Home | History | Annotate | Line # | Download | only in INSTALL
      1  1.20    mrg <!DOCTYPE html>
      2  1.10    mrg <html>
      3  1.20    mrg <!-- Created by GNU Texinfo 7.2, https://www.gnu.org/software/texinfo/ -->
      4  1.20    mrg <head>
      5  1.20    mrg <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
      6  1.20    mrg <!-- Copyright  1988-2024 Free Software Foundation, Inc.
      7   1.1    mrg 
      8  1.10    mrg Permission is granted to copy, distribute and/or modify this document
      9   1.3  skrll under the terms of the GNU Free Documentation License, Version 1.3 or
     10   1.1    mrg any later version published by the Free Software Foundation; with no
     11   1.1    mrg Invariant Sections, the Front-Cover texts being (a) (see below), and
     12   1.1    mrg with the Back-Cover Texts being (b) (see below).  A copy of the
     13  1.10    mrg license is included in the section entitled "GNU
     14  1.10    mrg Free Documentation License".
     15   1.1    mrg 
     16   1.1    mrg (a) The FSF's Front-Cover Text is:
     17   1.1    mrg 
     18  1.10    mrg A GNU Manual
     19   1.1    mrg 
     20   1.1    mrg (b) The FSF's Back-Cover Text is:
     21   1.1    mrg 
     22  1.10    mrg You have freedom to copy and modify this GNU Manual, like GNU
     23   1.1    mrg      software.  Copies published by the Free Software Foundation raise
     24  1.10    mrg      funds for GNU development. -->
     25  1.13    mrg <title>Installing GCC: Configuration</title>
     26  1.10    mrg 
     27  1.13    mrg <meta name="description" content="Installing GCC: Configuration">
     28  1.13    mrg <meta name="keywords" content="Installing GCC: Configuration">
     29  1.10    mrg <meta name="resource-type" content="document">
     30  1.10    mrg <meta name="distribution" content="global">
     31  1.10    mrg <meta name="Generator" content="makeinfo">
     32  1.20    mrg <meta name="viewport" content="width=device-width,initial-scale=1">
     33  1.20    mrg 
     34  1.10    mrg <style type="text/css">
     35  1.10    mrg <!--
     36  1.20    mrg a.copiable-link {visibility: hidden; text-decoration: none; line-height: 0em}
     37  1.10    mrg div.example {margin-left: 3.2em}
     38  1.20    mrg span:hover a.copiable-link {visibility: visible}
     39  1.20    mrg ul.mark-bullet {list-style-type: disc}
     40  1.20    mrg ul.mark-minus {list-style-type: "\2212"}
     41  1.10    mrg -->
     42  1.10    mrg </style>
     43  1.10    mrg 
     44  1.10    mrg 
     45   1.1    mrg </head>
     46  1.10    mrg 
     47  1.10    mrg <body lang="en">
     48  1.10    mrg 
     49  1.10    mrg 
     50  1.10    mrg 
     51  1.10    mrg 
     52  1.10    mrg 
     53  1.10    mrg 
     54  1.10    mrg 
     55  1.10    mrg 
     56  1.10    mrg 
     57  1.10    mrg 
     58  1.10    mrg 
     59  1.10    mrg 
     60  1.10    mrg 
     61  1.10    mrg 
     62  1.10    mrg 
     63  1.10    mrg 
     64  1.10    mrg 
     65  1.10    mrg 
     66  1.10    mrg 
     67  1.19    mrg 
     68  1.19    mrg 
     69  1.19    mrg 
     70  1.20    mrg <a class="index-entry-id" id="index-Configuration"></a>
     71  1.20    mrg <a class="index-entry-id" id="index-Installing-GCC_003a-Configuration"></a>
     72  1.10    mrg 
     73  1.10    mrg <p>Like most GNU software, GCC must be configured before it can be built.
     74   1.9    mrg This document describes the recommended configuration procedure
     75   1.9    mrg for both native and cross targets.
     76  1.10    mrg </p>
     77  1.20    mrg <p>We use <var class="var">srcdir</var> to refer to the toplevel source directory for
     78  1.20    mrg GCC; we use <var class="var">objdir</var> to refer to the toplevel build/object directory.
     79  1.10    mrg </p>
     80  1.20    mrg <p>If you obtained the sources by cloning the repository, <var class="var">srcdir</var>
     81  1.20    mrg must refer to the top <samp class="file">gcc</samp> directory, the one where the
     82  1.20    mrg <samp class="file">MAINTAINERS</samp> file can be found, and not its <samp class="file">gcc</samp>
     83  1.11    mrg subdirectory, otherwise the build will fail.
     84  1.10    mrg </p>
     85  1.20    mrg <p>If either <var class="var">srcdir</var> or <var class="var">objdir</var> is located on an automounted NFS
     86  1.20    mrg file system, the shell&rsquo;s built-in <code class="command">pwd</code> command will return
     87   1.1    mrg temporary pathnames.  Using these can lead to various sorts of build
     88  1.20    mrg problems.  To avoid this issue, set the <code class="env">PWDCMD</code> environment
     89  1.20    mrg variable to an automounter-aware <code class="command">pwd</code> command, e.g.,
     90  1.20    mrg <code class="command">pawd</code> or &lsquo;<samp class="samp">amq -w</samp>&rsquo;, during the configuration and build
     91   1.1    mrg phases.
     92  1.10    mrg </p>
     93  1.20    mrg <p>First, we <strong class="strong">highly</strong> recommend that GCC be built into a
     94  1.20    mrg separate directory from the sources which does <strong class="strong">not</strong> reside
     95   1.1    mrg within the source tree.  This is how we generally build GCC; building
     96  1.20    mrg where <var class="var">objdir</var> is a subdirectory of <var class="var">srcdir</var> should work as well;
     97  1.20    mrg building where <var class="var">objdir</var> == <var class="var">srcdir</var> is unsupported.
     98  1.10    mrg </p>
     99  1.10    mrg <p>If you have previously built GCC in the same directory for a
    100  1.20    mrg different target machine, do &lsquo;<samp class="samp">make distclean</samp>&rsquo; to delete all files
    101  1.20    mrg that might be invalid.  One of the files this deletes is <samp class="file">Makefile</samp>;
    102  1.20    mrg if &lsquo;<samp class="samp">make distclean</samp>&rsquo; complains that <samp class="file">Makefile</samp> does not exist
    103  1.10    mrg or issues a message like &ldquo;don&rsquo;t know how to make distclean&rdquo; it probably
    104   1.1    mrg means that the directory is already suitably clean.  However, with the
    105  1.20    mrg recommended method of building in a separate <var class="var">objdir</var>, you should
    106  1.20    mrg simply use a different <var class="var">objdir</var> for each target.
    107  1.10    mrg </p>
    108  1.20    mrg <p>Second, when configuring a native system, either <code class="command">cc</code> or
    109  1.20    mrg <code class="command">gcc</code> must be in your path or you must set <code class="env">CC</code> in
    110   1.1    mrg your environment before running configure.  Otherwise the configuration
    111   1.1    mrg scripts may fail.
    112  1.10    mrg </p>
    113   1.1    mrg 
    114  1.10    mrg <p>To configure GCC:
    115  1.10    mrg </p>
    116  1.20    mrg <div class="example smallexample">
    117  1.20    mrg <pre class="example-preformatted">% mkdir <var class="var">objdir</var>
    118  1.20    mrg % cd <var class="var">objdir</var>
    119  1.20    mrg % <var class="var">srcdir</var>/configure [<var class="var">options</var>] [<var class="var">target</var>]
    120  1.10    mrg </pre></div>
    121   1.1    mrg 
    122  1.20    mrg <h3 class="heading" id="Distributor-options"><span>Distributor options<a class="copiable-link" href="#Distributor-options"> &para;</a></span></h3>
    123   1.1    mrg 
    124   1.1    mrg <p>If you will be distributing binary versions of GCC, with modifications
    125   1.1    mrg to the source code, you should use the options described in this
    126   1.1    mrg section to make clear that your version contains modifications.
    127  1.10    mrg </p>
    128  1.20    mrg <dl class="table">
    129  1.20    mrg <dt><code class="code">--with-pkgversion=<var class="var">version</var></code></dt>
    130  1.10    mrg <dd><p>Specify a string that identifies your package.  You may wish
    131   1.1    mrg to include a build number or build date.  This version string will be
    132  1.20    mrg included in the output of <code class="command">gcc --version</code>.  This suffix does
    133  1.20    mrg not replace the default version string, only the &lsquo;<samp class="samp">GCC</samp>&rsquo; part.
    134  1.10    mrg </p>
    135  1.20    mrg <p>The default value is &lsquo;<samp class="samp">GCC</samp>&rsquo;.
    136  1.10    mrg </p>
    137  1.10    mrg </dd>
    138  1.20    mrg <dt><code class="code">--with-bugurl=<var class="var">url</var></code></dt>
    139  1.10    mrg <dd><p>Specify the URL that users should visit if they wish to report a bug.
    140   1.1    mrg You are of course welcome to forward bugs reported to you to the FSF,
    141   1.1    mrg if you determine that they are not bugs in your modifications.
    142  1.10    mrg </p>
    143  1.10    mrg <p>The default value refers to the FSF&rsquo;s GCC bug tracker.
    144  1.10    mrg </p>
    145  1.10    mrg </dd>
    146  1.20    mrg <dt><code class="code">--with-documentation-root-url=<var class="var">url</var></code></dt>
    147  1.20    mrg <dd><p>Specify the URL root that contains GCC option documentation.  The <var class="var">url</var>
    148  1.20    mrg should end with a <code class="code">/</code> character.
    149  1.13    mrg </p>
    150  1.20    mrg <p>The default value is <a class="uref" href="https://gcc.gnu.org/onlinedocs/">https://gcc.gnu.org/onlinedocs/</a>
    151  1.19    mrg on the GCC main development trunk.  On release branches, the default
    152  1.20    mrg is <code class="code">https://gcc.gnu.org/onlinedocs/gcc-<var class="var">major</var>.<var class="var">minor</var>.0/</code>.
    153  1.13    mrg </p>
    154  1.13    mrg </dd>
    155  1.20    mrg <dt><code class="code">--with-changes-root-url=<var class="var">url</var></code></dt>
    156  1.13    mrg <dd><p>Specify the URL root that contains information about changes in GCC
    157  1.20    mrg releases like <code class="code">gcc-<var class="var">version</var>/changes.html</code>.
    158  1.20    mrg The <var class="var">url</var> should end with a <code class="code">/</code> character.
    159  1.13    mrg </p>
    160  1.20    mrg <p>The default value is <a class="uref" href="https://gcc.gnu.org/">https://gcc.gnu.org/</a>.
    161  1.13    mrg </p>
    162  1.13    mrg </dd>
    163  1.10    mrg </dl>
    164   1.1    mrg 
    165  1.20    mrg <h3 class="heading" id="Host_002c-Build-and-Target-specification"><span>Host, Build and Target specification<a class="copiable-link" href="#Host_002c-Build-and-Target-specification"> &para;</a></span></h3>
    166   1.1    mrg 
    167  1.16    mrg <p>Specify the host, build and target machine configurations.  You do this
    168  1.20    mrg when you run the <samp class="file">configure</samp> script.
    169  1.16    mrg </p>
    170  1.20    mrg <p>The <em class="dfn">build</em> machine is the system which you are using, the
    171  1.20    mrg <em class="dfn">host</em> machine is the system where you want to run the resulting
    172  1.20    mrg compiler (normally the build machine), and the <em class="dfn">target</em> machine is
    173  1.16    mrg the system for which you want the compiler to generate code.
    174  1.16    mrg </p>
    175  1.16    mrg <p>If you are building a compiler to produce code for the machine it runs
    176  1.16    mrg on (a native compiler), you normally do not need to specify any operands
    177  1.20    mrg to <samp class="file">configure</samp>; it will try to guess the type of machine you are on
    178  1.16    mrg and use that as the build, host and target machines.  So you don&rsquo;t need
    179  1.16    mrg to specify a configuration when building a native compiler unless
    180  1.20    mrg <samp class="file">configure</samp> cannot figure out what your configuration is or guesses
    181  1.16    mrg wrong.
    182  1.16    mrg </p>
    183  1.20    mrg <p>In those cases, specify the build machine&rsquo;s <em class="dfn">configuration name</em>
    184  1.20    mrg with the <samp class="option">--host</samp> option; the host and target will default to be
    185  1.16    mrg the same as the host machine.
    186  1.16    mrg </p>
    187  1.16    mrg <p>Here is an example:
    188  1.16    mrg </p>
    189  1.20    mrg <div class="example smallexample">
    190  1.20    mrg <pre class="example-preformatted">./configure --host=x86_64-pc-linux-gnu
    191  1.16    mrg </pre></div>
    192  1.10    mrg 
    193  1.16    mrg <p>A configuration name may be canonical or it may be more or less
    194  1.20    mrg abbreviated (<samp class="file">config.sub</samp> script produces canonical versions).
    195  1.16    mrg </p>
    196  1.16    mrg <p>A canonical configuration name has three parts, separated by dashes.
    197  1.20    mrg It looks like this: &lsquo;<samp class="samp"><var class="var">cpu</var>-<var class="var">company</var>-<var class="var">system</var></samp>&rsquo;.
    198  1.16    mrg </p>
    199  1.16    mrg <p>Here are the possible CPU types:
    200  1.16    mrg </p>
    201  1.20    mrg <blockquote class="quotation">
    202  1.16    mrg <p>aarch64, aarch64_be, alpha, alpha64, amdgcn, arc, arceb, arm, armeb, avr, bfin,
    203  1.19    mrg bpf, cris, csky, epiphany, fido, fr30, frv, ft32, h8300, hppa, hppa2.0,
    204  1.16    mrg hppa64, i486, i686, ia64, iq2000, lm32, loongarch64, m32c, m32r, m32rle, m68k,
    205  1.16    mrg mcore, microblaze, microblazeel, mips, mips64, mips64el, mips64octeon,
    206  1.16    mrg mips64orion, mips64vr, mipsel, mipsisa32, mipsisa32r2, mipsisa64, mipsisa64r2,
    207  1.16    mrg mipsisa64r2el, mipsisa64sb1, mipsisa64sr71k, mipstx39, mmix, mn10300, moxie,
    208  1.16    mrg msp430, nds32be, nds32le, nios2, nvptx, or1k, pdp11, powerpc, powerpc64,
    209  1.16    mrg powerpc64le, powerpcle, pru, riscv32, riscv32be, riscv64, riscv64be, rl78, rx,
    210  1.19    mrg s390, s390x, sh, shle, sparc, sparc64, tic6x, v850,
    211  1.16    mrg v850e, v850e1, vax, visium, x86_64, xstormy16, xtensa
    212  1.16    mrg </p></blockquote>
    213  1.16    mrg 
    214  1.16    mrg <p>Here is a list of system types:
    215  1.16    mrg </p>
    216  1.20    mrg <blockquote class="quotation">
    217  1.20    mrg <p>aix<var class="var">version</var>, amdhsa, aout, cygwin, darwin<var class="var">version</var>,
    218  1.16    mrg eabi, eabialtivec, eabisim, eabisimaltivec, elf, elf32,
    219  1.20    mrg elfbare, elfoabi, freebsd<var class="var">version</var>, gnu, hpux, hpux<var class="var">version</var>,
    220  1.16    mrg kfreebsd-gnu, kopensolaris-gnu, linux-androideabi, linux-gnu,
    221  1.16    mrg linux-gnu_altivec, linux-musl, linux-uclibc, lynxos, mingw32, mingw32crt,
    222  1.20    mrg mmixware, msdosdjgpp, netbsd, netbsdelf<var class="var">version</var>, nto-qnx, openbsd,
    223  1.20    mrg rtems, solaris<var class="var">version</var>, symbianelf, tpf, uclinux, uclinux_eabi, vms,
    224  1.16    mrg vxworks, vxworksae, vxworksmils
    225  1.16    mrg </p></blockquote>
    226   1.1    mrg 
    227  1.20    mrg <h3 class="heading" id="Options-specification"><span>Options specification<a class="copiable-link" href="#Options-specification"> &para;</a></span></h3>
    228   1.1    mrg 
    229  1.20    mrg <p>Use <var class="var">options</var> to override several configure time options for
    230  1.20    mrg GCC.  A list of supported <var class="var">options</var> follows; &lsquo;<samp class="samp">configure
    231  1.10    mrg --help</samp>&rsquo; may list other options, but those not listed below may not
    232   1.1    mrg work and should not normally be used.
    233  1.10    mrg </p>
    234  1.20    mrg <p>Note that each <samp class="option">--enable</samp> option has a corresponding
    235  1.20    mrg <samp class="option">--disable</samp> option and that each <samp class="option">--with</samp> option has a
    236  1.20    mrg corresponding <samp class="option">--without</samp> option.
    237  1.10    mrg </p>
    238  1.20    mrg <dl class="table">
    239  1.20    mrg <dt><code class="code">--prefix=<var class="var">dirname</var></code></dt>
    240  1.10    mrg <dd><p>Specify the toplevel installation
    241   1.1    mrg directory.  This is the recommended way to install the tools into a directory
    242   1.1    mrg other than the default.  The toplevel installation directory defaults to
    243  1.20    mrg <samp class="file">/usr/local</samp>.
    244  1.10    mrg </p>
    245  1.20    mrg <p>We <strong class="strong">highly</strong> recommend against <var class="var">dirname</var> being the same or a
    246  1.20    mrg subdirectory of <var class="var">objdir</var> or vice versa.  If specifying a directory
    247  1.10    mrg beneath a user&rsquo;s home directory tree, some shells will not expand
    248  1.20    mrg <var class="var">dirname</var> correctly if it contains the &lsquo;<samp class="samp">~</samp>&rsquo; metacharacter; use
    249  1.20    mrg <code class="env">$HOME</code> instead.
    250  1.10    mrg </p>
    251  1.20    mrg <p>The following standard <code class="command">autoconf</code> options are supported.  Normally you
    252   1.1    mrg should not need to use these options.
    253  1.20    mrg </p><dl class="table">
    254  1.20    mrg <dt><code class="code">--exec-prefix=<var class="var">dirname</var></code></dt>
    255  1.10    mrg <dd><p>Specify the toplevel installation directory for architecture-dependent
    256  1.20    mrg files.  The default is <samp class="file"><var class="var">prefix</var></samp>.
    257  1.10    mrg </p>
    258  1.10    mrg </dd>
    259  1.20    mrg <dt><code class="code">--bindir=<var class="var">dirname</var></code></dt>
    260  1.10    mrg <dd><p>Specify the installation directory for the executables called by users
    261  1.20    mrg (such as <code class="command">gcc</code> and <code class="command">g++</code>).  The default is
    262  1.20    mrg <samp class="file"><var class="var">exec-prefix</var>/bin</samp>.
    263  1.10    mrg </p>
    264  1.10    mrg </dd>
    265  1.20    mrg <dt><code class="code">--libdir=<var class="var">dirname</var></code></dt>
    266  1.10    mrg <dd><p>Specify the installation directory for object code libraries and
    267  1.20    mrg internal data files of GCC.  The default is <samp class="file"><var class="var">exec-prefix</var>/lib</samp>.
    268  1.10    mrg </p>
    269  1.10    mrg </dd>
    270  1.20    mrg <dt><code class="code">--libexecdir=<var class="var">dirname</var></code></dt>
    271  1.10    mrg <dd><p>Specify the installation directory for internal executables of GCC.
    272  1.20    mrg The default is <samp class="file"><var class="var">exec-prefix</var>/libexec</samp>.
    273  1.10    mrg </p>
    274  1.10    mrg </dd>
    275  1.20    mrg <dt><code class="code">--with-slibdir=<var class="var">dirname</var></code></dt>
    276  1.10    mrg <dd><p>Specify the installation directory for the shared libgcc library.  The
    277  1.20    mrg default is <samp class="file"><var class="var">libdir</var></samp>.
    278  1.10    mrg </p>
    279  1.10    mrg </dd>
    280  1.20    mrg <dt><code class="code">--datarootdir=<var class="var">dirname</var></code></dt>
    281  1.10    mrg <dd><p>Specify the root of the directory tree for read-only architecture-independent
    282  1.20    mrg data files referenced by GCC.  The default is <samp class="file"><var class="var">prefix</var>/share</samp>.
    283  1.10    mrg </p>
    284  1.10    mrg </dd>
    285  1.20    mrg <dt><code class="code">--infodir=<var class="var">dirname</var></code></dt>
    286  1.10    mrg <dd><p>Specify the installation directory for documentation in info format.
    287  1.20    mrg The default is <samp class="file"><var class="var">datarootdir</var>/info</samp>.
    288  1.10    mrg </p>
    289  1.10    mrg </dd>
    290  1.20    mrg <dt><code class="code">--datadir=<var class="var">dirname</var></code></dt>
    291  1.10    mrg <dd><p>Specify the installation directory for some architecture-independent
    292  1.20    mrg data files referenced by GCC.  The default is <samp class="file"><var class="var">datarootdir</var></samp>.
    293  1.10    mrg </p>
    294  1.10    mrg </dd>
    295  1.20    mrg <dt><code class="code">--docdir=<var class="var">dirname</var></code></dt>
    296  1.10    mrg <dd><p>Specify the installation directory for documentation files (other
    297  1.20    mrg than Info) for GCC.  The default is <samp class="file"><var class="var">datarootdir</var>/doc</samp>.
    298  1.10    mrg </p>
    299  1.10    mrg </dd>
    300  1.20    mrg <dt><code class="code">--htmldir=<var class="var">dirname</var></code></dt>
    301  1.10    mrg <dd><p>Specify the installation directory for HTML documentation files.
    302  1.20    mrg The default is <samp class="file"><var class="var">docdir</var></samp>.
    303  1.10    mrg </p>
    304  1.10    mrg </dd>
    305  1.20    mrg <dt><code class="code">--pdfdir=<var class="var">dirname</var></code></dt>
    306  1.10    mrg <dd><p>Specify the installation directory for PDF documentation files.
    307  1.20    mrg The default is <samp class="file"><var class="var">docdir</var></samp>.
    308  1.10    mrg </p>
    309  1.10    mrg </dd>
    310  1.20    mrg <dt><code class="code">--mandir=<var class="var">dirname</var></code></dt>
    311  1.10    mrg <dd><p>Specify the installation directory for manual pages.  The default is
    312  1.20    mrg <samp class="file"><var class="var">datarootdir</var>/man</samp>.  (Note that the manual pages are only extracts
    313   1.1    mrg from the full GCC manuals, which are provided in Texinfo format.  The manpages
    314   1.1    mrg are derived by an automatic conversion process from parts of the full
    315   1.1    mrg manual.)
    316  1.10    mrg </p>
    317  1.10    mrg </dd>
    318  1.20    mrg <dt><code class="code">--with-gxx-include-dir=<var class="var">dirname</var></code></dt>
    319  1.10    mrg <dd><p>Specify
    320   1.1    mrg the installation directory for G++ header files.  The default depends
    321   1.1    mrg on other configuration options, and differs between cross and native
    322   1.1    mrg configurations.
    323  1.10    mrg </p>
    324  1.10    mrg </dd>
    325  1.20    mrg <dt><code class="code">--with-specs=<var class="var">specs</var></code></dt>
    326  1.10    mrg <dd><p>Specify additional command line driver SPECS.
    327   1.3  skrll This can be useful if you need to turn on a non-standard feature by
    328  1.10    mrg default without modifying the compiler&rsquo;s source code, for instance
    329  1.20    mrg <samp class="option">--with-specs=%{!fcommon:%{!fno-common:-fno-common}}</samp>.
    330   1.3  skrll See &ldquo;Spec Files&rdquo; in the main manual
    331  1.10    mrg </p>
    332  1.10    mrg </dd>
    333  1.10    mrg </dl>
    334   1.3  skrll 
    335  1.10    mrg </dd>
    336  1.20    mrg <dt><code class="code">--program-prefix=<var class="var">prefix</var></code></dt>
    337  1.10    mrg <dd><p>GCC supports some transformations of the names of its programs when
    338  1.20    mrg installing them.  This option prepends <var class="var">prefix</var> to the names of
    339  1.20    mrg programs to install in <var class="var">bindir</var> (see above).  For example, specifying
    340  1.20    mrg <samp class="option">--program-prefix=foo-</samp> would result in &lsquo;<samp class="samp">gcc</samp>&rsquo;
    341  1.20    mrg being installed as <samp class="file">/usr/local/bin/foo-gcc</samp>.
    342  1.20    mrg </p>
    343  1.20    mrg </dd>
    344  1.20    mrg <dt><code class="code">--program-suffix=<var class="var">suffix</var></code></dt>
    345  1.20    mrg <dd><p>Appends <var class="var">suffix</var> to the names of programs to install in <var class="var">bindir</var>
    346  1.20    mrg (see above).  For example, specifying <samp class="option">--program-suffix=-3.1</samp>
    347  1.20    mrg would result in &lsquo;<samp class="samp">gcc</samp>&rsquo; being installed as
    348  1.20    mrg <samp class="file">/usr/local/bin/gcc-3.1</samp>.
    349  1.20    mrg </p>
    350  1.20    mrg </dd>
    351  1.20    mrg <dt><code class="code">--program-transform-name=<var class="var">pattern</var></code></dt>
    352  1.20    mrg <dd><p>Applies the &lsquo;<samp class="samp">sed</samp>&rsquo; script <var class="var">pattern</var> to be applied to the names
    353  1.20    mrg of programs to install in <var class="var">bindir</var> (see above).  <var class="var">pattern</var> has to
    354  1.20    mrg consist of one or more basic &lsquo;<samp class="samp">sed</samp>&rsquo; editing commands, separated by
    355  1.20    mrg semicolons.  For example, if you want the &lsquo;<samp class="samp">gcc</samp>&rsquo; program name to be
    356  1.20    mrg transformed to the installed program <samp class="file">/usr/local/bin/myowngcc</samp> and
    357  1.20    mrg the &lsquo;<samp class="samp">g++</samp>&rsquo; program name to be transformed to
    358  1.20    mrg <samp class="file">/usr/local/bin/gspecial++</samp> without changing other program names,
    359   1.1    mrg you could use the pattern
    360  1.20    mrg <samp class="option">--program-transform-name='s/^gcc$/myowngcc/; s/^g++$/gspecial++/'</samp>
    361   1.1    mrg to achieve this effect.
    362  1.10    mrg </p>
    363  1.10    mrg <p>All three options can be combined and used together, resulting in more
    364  1.20    mrg complex conversion patterns.  As a basic rule, <var class="var">prefix</var> (and
    365  1.20    mrg <var class="var">suffix</var>) are prepended (appended) before further transformations
    366  1.20    mrg can happen with a special transformation script <var class="var">pattern</var>.
    367  1.10    mrg </p>
    368  1.10    mrg <p>As currently implemented, this option only takes effect for native
    369  1.10    mrg builds; cross compiler binaries&rsquo; names are not transformed even when a
    370   1.1    mrg transformation is explicitly asked for by one of these options.
    371  1.10    mrg </p>
    372  1.10    mrg <p>For native builds, some of the installed programs are also installed
    373   1.1    mrg with the target alias in front of their name, as in
    374  1.20    mrg &lsquo;<samp class="samp">i686-pc-linux-gnu-gcc</samp>&rsquo;.  All of the above transformations happen
    375   1.1    mrg before the target alias is prepended to the name&mdash;so, specifying
    376  1.20    mrg <samp class="option">--program-prefix=foo-</samp> and <samp class="option">program-suffix=-3.1</samp>, the
    377   1.1    mrg resulting binary would be installed as
    378  1.20    mrg <samp class="file">/usr/local/bin/i686-pc-linux-gnu-foo-gcc-3.1</samp>.
    379  1.10    mrg </p>
    380  1.10    mrg <p>As a last shortcoming, none of the installed Ada programs are
    381   1.1    mrg transformed yet, which will be fixed in some time.
    382  1.10    mrg </p>
    383  1.10    mrg </dd>
    384  1.20    mrg <dt><code class="code">--with-local-prefix=<var class="var">dirname</var></code></dt>
    385  1.10    mrg <dd><p>Specify the
    386   1.1    mrg installation directory for local include files.  The default is
    387  1.20    mrg <samp class="file">/usr/local</samp>.  Specify this option if you want the compiler to
    388  1.20    mrg search directory <samp class="file"><var class="var">dirname</var>/include</samp> for locally installed
    389  1.20    mrg header files <em class="emph">instead</em> of <samp class="file">/usr/local/include</samp>.
    390  1.10    mrg </p>
    391  1.20    mrg <p>You should specify <samp class="option">--with-local-prefix</samp> <strong class="strong">only</strong> if your
    392  1.20    mrg site has a different convention (not <samp class="file">/usr/local</samp>) for where to put
    393   1.1    mrg site-specific files.
    394  1.10    mrg </p>
    395  1.20    mrg <p>The default value for <samp class="option">--with-local-prefix</samp> is <samp class="file">/usr/local</samp>
    396  1.20    mrg regardless of the value of <samp class="option">--prefix</samp>.  Specifying
    397  1.20    mrg <samp class="option">--prefix</samp> has no effect on which directory GCC searches for
    398   1.1    mrg local header files.  This may seem counterintuitive, but actually it is
    399   1.1    mrg logical.
    400  1.10    mrg </p>
    401  1.20    mrg <p>The purpose of <samp class="option">--prefix</samp> is to specify where to <em class="emph">install
    402  1.20    mrg GCC</em>.  The local header files in <samp class="file">/usr/local/include</samp>&mdash;if you put
    403   1.1    mrg any in that directory&mdash;are not part of GCC.  They are part of other
    404   1.1    mrg programs&mdash;perhaps many others.  (GCC installs its own header files in
    405  1.20    mrg another directory which is based on the <samp class="option">--prefix</samp> value.)
    406  1.10    mrg </p>
    407  1.10    mrg <p>Both the local-prefix include directory and the GCC-prefix include
    408  1.10    mrg directory are part of GCC&rsquo;s &ldquo;system include&rdquo; directories.  Although these
    409   1.1    mrg two directories are not fixed, they need to be searched in the proper
    410   1.1    mrg order for the correct processing of the include_next directive.  The
    411   1.1    mrg local-prefix include directory is searched before the GCC-prefix
    412   1.1    mrg include directory.  Another characteristic of system include directories
    413   1.1    mrg is that pedantic warnings are turned off for headers in these directories.
    414  1.10    mrg </p>
    415  1.20    mrg <p>Some autoconf macros add <samp class="option">-I <var class="var">directory</var></samp> options to the
    416   1.1    mrg compiler command line, to ensure that directories containing installed
    417  1.20    mrg packages&rsquo; headers are searched.  When <var class="var">directory</var> is one of GCC&rsquo;s
    418   1.1    mrg system include directories, GCC will ignore the option so that system
    419   1.1    mrg directories continue to be processed in the correct order.  This
    420   1.1    mrg may result in a search order different from what was specified but the
    421   1.1    mrg directory will still be searched.
    422  1.10    mrg </p>
    423  1.10    mrg <p>GCC automatically searches for ordinary libraries using
    424  1.20    mrg <code class="env">GCC_EXEC_PREFIX</code>.  Thus, when the same installation prefix is
    425   1.1    mrg used for both GCC and packages, GCC will automatically search for
    426   1.1    mrg both headers and libraries.  This provides a configuration that is
    427   1.1    mrg easy to use.  GCC behaves in a manner similar to that when it is
    428  1.20    mrg installed as a system compiler in <samp class="file">/usr</samp>.
    429  1.10    mrg </p>
    430  1.10    mrg <p>Sites that need to install multiple versions of GCC may not want to
    431   1.1    mrg use the above simple configuration.  It is possible to use the
    432  1.20    mrg <samp class="option">--program-prefix</samp>, <samp class="option">--program-suffix</samp> and
    433  1.20    mrg <samp class="option">--program-transform-name</samp> options to install multiple versions
    434   1.1    mrg into a single directory, but it may be simpler to use different prefixes
    435  1.20    mrg and the <samp class="option">--with-local-prefix</samp> option to specify the location of the
    436   1.1    mrg site-specific files for each version.  It will then be necessary for
    437   1.1    mrg users to specify explicitly the location of local site libraries
    438  1.20    mrg (e.g., with <code class="env">LIBRARY_PATH</code>).
    439  1.10    mrg </p>
    440  1.20    mrg <p>The same value can be used for both <samp class="option">--with-local-prefix</samp> and
    441  1.20    mrg <samp class="option">--prefix</samp> provided it is not <samp class="file">/usr</samp>.  This can be used
    442  1.20    mrg to avoid the default search of <samp class="file">/usr/local/include</samp>.
    443  1.10    mrg </p>
    444  1.20    mrg <p><strong class="strong">Do not</strong> specify <samp class="file">/usr</samp> as the <samp class="option">--with-local-prefix</samp>!
    445  1.20    mrg The directory you use for <samp class="option">--with-local-prefix</samp> <strong class="strong">must not</strong>
    446  1.10    mrg contain any of the system&rsquo;s standard header files.  If it did contain
    447   1.1    mrg them, certain programs would be miscompiled (including GNU Emacs, on
    448   1.1    mrg certain targets), because this would override and nullify the header
    449  1.20    mrg file corrections made by the <code class="command">fixincludes</code> script.
    450  1.10    mrg </p>
    451  1.10    mrg <p>Indications are that people who use this option use it based on mistaken
    452   1.1    mrg ideas of what it is for.  People use it as if it specified where to
    453   1.1    mrg install part of GCC.  Perhaps they make this assumption because
    454   1.1    mrg installing GCC creates the directory.
    455  1.10    mrg </p>
    456  1.10    mrg </dd>
    457  1.20    mrg <dt><code class="code">--with-gcc-major-version-only</code></dt>
    458  1.10    mrg <dd><p>Specifies that GCC should use only the major number rather than
    459  1.20    mrg <var class="var">major</var>.<var class="var">minor</var>.<var class="var">patchlevel</var> in filesystem paths.
    460  1.10    mrg </p>
    461  1.10    mrg </dd>
    462  1.20    mrg <dt><code class="code">--with-native-system-header-dir=<var class="var">dirname</var></code></dt>
    463  1.20    mrg <dd><p>Specifies that <var class="var">dirname</var> is the directory that contains native system
    464  1.20    mrg header files, rather than <samp class="file">/usr/include</samp>.  This option is most useful
    465   1.3  skrll if you are creating a compiler that should be isolated from the system
    466   1.3  skrll as much as possible.  It is most commonly used with the
    467  1.20    mrg <samp class="option">--with-sysroot</samp> option and will cause GCC to search
    468  1.20    mrg <var class="var">dirname</var> inside the system root specified by that option.
    469  1.10    mrg </p>
    470  1.10    mrg </dd>
    471  1.20    mrg <dt><code class="code">--enable-shared[=<var class="var">package</var>[,&hellip;]]</code></dt>
    472  1.10    mrg <dd><p>Build shared versions of libraries, if shared libraries are supported on
    473   1.1    mrg the target platform.  Unlike GCC 2.95.x and earlier, shared libraries
    474   1.1    mrg are enabled by default on all platforms that support shared libraries.
    475  1.10    mrg </p>
    476  1.10    mrg <p>If a list of packages is given as an argument, build shared libraries
    477   1.1    mrg only for the listed packages.  For other packages, only static libraries
    478   1.1    mrg will be built.  Package names currently recognized in the GCC tree are
    479  1.20    mrg &lsquo;<samp class="samp">libgcc</samp>&rsquo; (also known as &lsquo;<samp class="samp">gcc</samp>&rsquo;), &lsquo;<samp class="samp">libstdc++</samp>&rsquo; (not
    480  1.20    mrg &lsquo;<samp class="samp">libstdc++-v3</samp>&rsquo;), &lsquo;<samp class="samp">libffi</samp>&rsquo;, &lsquo;<samp class="samp">zlib</samp>&rsquo;, &lsquo;<samp class="samp">boehm-gc</samp>&rsquo;,
    481  1.20    mrg &lsquo;<samp class="samp">ada</samp>&rsquo;, &lsquo;<samp class="samp">libada</samp>&rsquo;, &lsquo;<samp class="samp">libgo</samp>&rsquo;, &lsquo;<samp class="samp">libobjc</samp>&rsquo;, and &lsquo;<samp class="samp">libphobos</samp>&rsquo;.
    482  1.20    mrg Note &lsquo;<samp class="samp">libiberty</samp>&rsquo; does not support shared libraries at all.
    483  1.20    mrg </p>
    484  1.20    mrg <p>Use <samp class="option">--disable-shared</samp> to build only static libraries.  Note that
    485  1.20    mrg <samp class="option">--disable-shared</samp> does not accept a list of package names as
    486  1.20    mrg argument, only <samp class="option">--enable-shared</samp> does.
    487  1.10    mrg </p>
    488  1.20    mrg <p>Contrast with <samp class="option">--enable-host-shared</samp>, which affects <em class="emph">host</em>
    489   1.5    mrg code.
    490  1.10    mrg </p>
    491  1.10    mrg </dd>
    492  1.20    mrg <dt><code class="code">--enable-host-shared</code></dt>
    493  1.20    mrg <dd><p>Specify that the <em class="emph">host</em> code should be built into position-independent
    494  1.20    mrg machine code (with <samp class="option">-fPIC</samp>), allowing it to be used within shared
    495  1.19    mrg libraries, but yielding a slightly slower compiler.
    496  1.10    mrg </p>
    497  1.10    mrg <p>This option is required when building the libgccjit.so library.
    498  1.10    mrg </p>
    499  1.20    mrg <p>Contrast with <samp class="option">--enable-shared</samp>, which affects <em class="emph">target</em>
    500   1.5    mrg libraries.
    501  1.10    mrg </p>
    502  1.10    mrg </dd>
    503  1.20    mrg <dt><code class="code">--enable-host-pie</code></dt>
    504  1.20    mrg <dd><p>Specify that the <em class="emph">host</em> executables should be built into
    505  1.20    mrg position-independent executables (with <samp class="option">-fPIE</samp> and <samp class="option">-pie</samp>),
    506  1.19    mrg yielding a slightly slower compiler (but faster than
    507  1.20    mrg <samp class="option">--enable-host-shared</samp>).  Position-independent executables are loaded
    508  1.19    mrg at random addresses each time they are executed, therefore provide additional
    509  1.19    mrg protection against Return Oriented Programming (ROP) attacks.
    510  1.19    mrg </p>
    511  1.20    mrg <p><samp class="option">--enable-host-pie</samp> may be used with <samp class="option">--enable-host-shared</samp>,
    512  1.20    mrg in which case <samp class="option">-fPIC</samp> is used when compiling, and <samp class="option">-pie</samp> when
    513  1.19    mrg linking.
    514  1.19    mrg </p>
    515  1.19    mrg </dd>
    516  1.20    mrg <dt><code class="code">--enable-host-bind-now</code></dt>
    517  1.20    mrg <dd><p>Specify that the <em class="emph">host</em> executables should be linked with the option
    518  1.20    mrg <samp class="option">-Wl,-z,now</samp>, which means that the dynamic linker will resolve all
    519  1.19    mrg symbols when the executables are started, and that in turn allows RELRO to
    520  1.19    mrg mark the GOT read-only, resulting in better security.
    521  1.19    mrg </p>
    522  1.19    mrg </dd>
    523  1.20    mrg <dt><code class="code"><a class="anchor" id="with-gnu-as"></a>--with-gnu-as</code></dt>
    524  1.10    mrg <dd><p>Specify that the compiler should assume that the
    525   1.1    mrg assembler it finds is the GNU assembler.  However, this does not modify
    526   1.1    mrg the rules to find an assembler and will result in confusion if the
    527   1.1    mrg assembler found is not actually the GNU assembler.  (Confusion may also
    528   1.1    mrg result if the compiler finds the GNU assembler but has not been
    529  1.20    mrg configured with <samp class="option">--with-gnu-as</samp>.)  If you have more than one
    530   1.1    mrg assembler installed on your system, you may want to use this option in
    531  1.20    mrg connection with <samp class="option">--with-as=<var class="var">pathname</var></samp> or
    532  1.20    mrg <samp class="option">--with-build-time-tools=<var class="var">pathname</var></samp>.
    533  1.10    mrg </p>
    534  1.10    mrg <p>The following systems are the only ones where it makes a difference
    535   1.1    mrg whether you use the GNU assembler.  On any other system,
    536  1.20    mrg <samp class="option">--with-gnu-as</samp> has no effect.
    537  1.10    mrg </p>
    538  1.20    mrg <ul class="itemize mark-bullet">
    539  1.20    mrg <li>&lsquo;<samp class="samp">hppa1.0-<var class="var">any</var>-<var class="var">any</var></samp>&rsquo;
    540  1.20    mrg </li><li>&lsquo;<samp class="samp">hppa1.1-<var class="var">any</var>-<var class="var">any</var></samp>&rsquo;
    541  1.20    mrg </li><li>&lsquo;<samp class="samp">*-*-solaris2.11</samp>&rsquo;
    542  1.10    mrg </li></ul>
    543  1.10    mrg 
    544  1.10    mrg </dd>
    545  1.20    mrg <dt><code class="code"><a class="anchor" id="with-as"></a>--with-as=<var class="var">pathname</var></code></dt>
    546  1.10    mrg <dd><p>Specify that the compiler should use the assembler pointed to by
    547  1.20    mrg <var class="var">pathname</var>, rather than the one found by the standard rules to find
    548   1.1    mrg an assembler, which are:
    549  1.20    mrg </p><ul class="itemize mark-bullet">
    550  1.20    mrg <li>Unless GCC is being built with a cross compiler, check the
    551  1.20    mrg <samp class="file"><var class="var">libexec</var>/gcc/<var class="var">target</var>/<var class="var">version</var></samp> directory.
    552  1.20    mrg <var class="var">libexec</var> defaults to <samp class="file"><var class="var">exec-prefix</var>/libexec</samp>;
    553  1.20    mrg <var class="var">exec-prefix</var> defaults to <var class="var">prefix</var>, which
    554  1.20    mrg defaults to <samp class="file">/usr/local</samp> unless overridden by the
    555  1.20    mrg <samp class="option">--prefix=<var class="var">pathname</var></samp> switch described above.  <var class="var">target</var>
    556  1.20    mrg is the target system triple, such as &lsquo;<samp class="samp">sparc-sun-solaris2.11</samp>&rsquo;, and
    557  1.20    mrg <var class="var">version</var> denotes the GCC version, such as 3.0.
    558   1.1    mrg 
    559  1.20    mrg </li><li>If the target system is the same that you are building on, check
    560  1.19    mrg operating system specific directories.
    561   1.1    mrg 
    562  1.20    mrg </li><li>Check in the <code class="env">PATH</code> for a tool whose name is prefixed by the
    563   1.1    mrg target system triple.
    564   1.1    mrg 
    565  1.20    mrg </li><li>Check in the <code class="env">PATH</code> for a tool whose name is not prefixed by the
    566   1.1    mrg target system triple, if the host and target system triple are
    567   1.1    mrg the same (in other words, we use a host tool if it can be used for
    568  1.10    mrg the target as well).
    569  1.10    mrg </li></ul>
    570   1.1    mrg 
    571  1.20    mrg <p>You may want to use <samp class="option">--with-as</samp> if no assembler
    572   1.1    mrg is installed in the directories listed above, or if you have multiple
    573   1.1    mrg assemblers installed and want to choose one that is not found by the
    574   1.1    mrg above rules.
    575  1.10    mrg </p>
    576  1.10    mrg </dd>
    577  1.20    mrg <dt><code class="code"><a class="anchor" id="with-gnu-ld"></a>--with-gnu-ld</code></dt>
    578  1.20    mrg <dd><p>Same as <a class="uref" href="#with-gnu-as"><samp class="option">--with-gnu-as</samp></a>
    579   1.1    mrg but for the linker.
    580  1.10    mrg </p>
    581  1.10    mrg </dd>
    582  1.20    mrg <dt><code class="code">--with-ld=<var class="var">pathname</var></code></dt>
    583  1.20    mrg <dd><p>Same as <a class="uref" href="#with-as"><samp class="option">--with-as</samp></a>
    584   1.1    mrg but for the linker.
    585  1.10    mrg </p>
    586  1.10    mrg </dd>
    587  1.20    mrg <dt><code class="code">--with-dsymutil=<var class="var">pathname</var></code></dt>
    588  1.20    mrg <dd><p>Same as <a class="uref" href="#with-as"><samp class="option">--with-as</samp></a>
    589  1.14    mrg but for the debug linker (only used on Darwin platforms so far).
    590  1.14    mrg </p>
    591  1.14    mrg </dd>
    592  1.20    mrg <dt><code class="code">--with-tls=<var class="var">dialect</var></code></dt>
    593  1.20    mrg <dd><p>Specify the default TLS dialect, for systems where there is a choice.
    594  1.20    mrg For ARM targets, possible values for <var class="var">dialect</var> are <code class="code">gnu</code> or
    595  1.20    mrg <code class="code">gnu2</code>, which select between the original GNU dialect and the GNU TLS
    596   1.3  skrll descriptor-based dialect.
    597  1.20    mrg For RISC-V targets, possible values for <var class="var">dialect</var> are <code class="code">trad</code> or
    598  1.20    mrg <code class="code">desc</code>, which select between the traditional GNU dialect and the GNU TLS
    599  1.19    mrg descriptor-based dialect.
    600  1.10    mrg </p>
    601  1.10    mrg </dd>
    602  1.20    mrg <dt><code class="code">--enable-multiarch</code></dt>
    603  1.10    mrg <dd><p>Specify whether to enable or disable multiarch support.  The default is
    604   1.3  skrll to check for glibc start files in a multiarch location, and enable it
    605   1.3  skrll if the files are found.  The auto detection is enabled for native builds,
    606  1.20    mrg and for cross builds configured with <samp class="option">--with-sysroot</samp>, and without
    607  1.20    mrg <samp class="option">--with-native-system-header-dir</samp>.
    608   1.3  skrll More documentation about multiarch can be found at
    609  1.20    mrg <a class="uref" href="https://wiki.debian.org/Multiarch">https://wiki.debian.org/Multiarch</a>.
    610  1.10    mrg </p>
    611  1.10    mrg </dd>
    612  1.20    mrg <dt><code class="code">--enable-sjlj-exceptions</code></dt>
    613  1.20    mrg <dd><p>Force use of the <code class="code">setjmp</code>/<code class="code">longjmp</code>-based scheme for exceptions.
    614  1.20    mrg &lsquo;<samp class="samp">configure</samp>&rsquo; ordinarily picks the correct value based on the platform.
    615   1.6    mrg Only use this option if you are sure you need a different setting.
    616  1.10    mrg </p>
    617  1.10    mrg </dd>
    618  1.20    mrg <dt><code class="code">--enable-vtable-verify</code></dt>
    619  1.10    mrg <dd><p>Specify whether to enable or disable the vtable verification feature.
    620   1.5    mrg Enabling this feature causes libstdc++ to be built with its virtual calls
    621   1.5    mrg in verifiable mode.  This means that, when linked with libvtv, every
    622   1.5    mrg virtual call in libstdc++ will verify the vtable pointer through which the
    623   1.5    mrg call will be made before actually making the call.  If not linked with libvtv,
    624  1.10    mrg the verifier will call stub functions (in libstdc++ itself) and do nothing.
    625   1.5    mrg If vtable verification is disabled, then libstdc++ is not built with its
    626   1.5    mrg virtual calls in verifiable mode at all.  However the libvtv library will
    627  1.20    mrg still be built (see <samp class="option">--disable-libvtv</samp> to turn off building libvtv).
    628  1.20    mrg <samp class="option">--disable-vtable-verify</samp> is the default.
    629  1.10    mrg </p>
    630  1.10    mrg </dd>
    631  1.20    mrg <dt><code class="code">--disable-gcov</code></dt>
    632  1.12    mrg <dd><p>Specify that the run-time library used for coverage analysis
    633  1.12    mrg and associated host tools should not be built.
    634  1.12    mrg </p>
    635  1.12    mrg </dd>
    636  1.20    mrg <dt><code class="code">--disable-multilib</code></dt>
    637  1.10    mrg <dd><p>Specify that multiple target
    638   1.1    mrg libraries to support different target variants, calling
    639   1.1    mrg conventions, etc. should not be built.  The default is to build a
    640   1.1    mrg predefined set of them.
    641  1.10    mrg </p>
    642  1.10    mrg <p>Some targets provide finer-grained control over which multilibs are built
    643  1.20    mrg (e.g., <samp class="option">--disable-softfloat</samp>):
    644  1.20    mrg </p><dl class="table">
    645  1.20    mrg <dt><code class="code">arm-*-*</code></dt>
    646  1.10    mrg <dd><p>fpu, 26bit, underscore, interwork, biendian, nofmult.
    647  1.10    mrg </p>
    648  1.10    mrg </dd>
    649  1.20    mrg <dt><code class="code">m68*-*-*</code></dt>
    650  1.10    mrg <dd><p>softfloat, m68881, m68000, m68020.
    651  1.10    mrg </p>
    652  1.10    mrg </dd>
    653  1.20    mrg <dt><code class="code">mips*-*-*</code></dt>
    654  1.10    mrg <dd><p>single-float, biendian, softfloat.
    655  1.10    mrg </p>
    656  1.10    mrg </dd>
    657  1.20    mrg <dt><code class="code">msp430-*-*</code></dt>
    658  1.13    mrg <dd><p>no-exceptions
    659  1.13    mrg </p>
    660  1.13    mrg </dd>
    661  1.20    mrg <dt><code class="code">powerpc*-*-*, rs6000*-*-*</code></dt>
    662  1.10    mrg <dd><p>aix64, pthread, softfloat, powercpu, powerpccpu, powerpcos, biendian,
    663   1.1    mrg sysv, aix.
    664  1.10    mrg </p>
    665  1.10    mrg </dd>
    666  1.10    mrg </dl>
    667   1.1    mrg 
    668  1.10    mrg </dd>
    669  1.20    mrg <dt><code class="code">--with-multilib-list=<var class="var">list</var></code></dt>
    670  1.20    mrg <dt><code class="code">--without-multilib-list</code></dt>
    671  1.20    mrg <dd><p>Specify what multilibs to build.  <var class="var">list</var> is a comma separated list of
    672   1.9    mrg values, possibly consisting of a single value.  Currently only implemented
    673  1.19    mrg for aarch64*-*-*, amdgcn*-*-*, arm*-*-*, loongarch*-*-*, riscv*-*-*, sh*-*-*
    674  1.19    mrg and x86-64-*-linux*.  The accepted values and meaning for each target is given
    675  1.16    mrg below.
    676  1.10    mrg </p>
    677  1.20    mrg <dl class="table">
    678  1.20    mrg <dt><code class="code">aarch64*-*-*</code></dt>
    679  1.20    mrg <dd><p><var class="var">list</var> is a comma separated list of <code class="code">ilp32</code>, and <code class="code">lp64</code>
    680  1.10    mrg to enable ILP32 and LP64 run-time libraries, respectively.  If
    681  1.20    mrg <var class="var">list</var> is empty, then there will be no multilibs and only the
    682  1.20    mrg default run-time library will be built.  If <var class="var">list</var> is
    683  1.20    mrg <code class="code">default</code> or &ndash;with-multilib-list= is not specified, then the
    684  1.10    mrg default set of libraries is selected based on the value of
    685  1.20    mrg <samp class="option">--target</samp>.
    686  1.10    mrg </p>
    687  1.10    mrg </dd>
    688  1.20    mrg <dt><code class="code">amdgcn*-*-*</code></dt>
    689  1.20    mrg <dd><p><var class="var">list</var> is a comma separated list of ISA names (allowed values: <code class="code">fiji</code>,
    690  1.20    mrg <code class="code">gfx900</code>, <code class="code">gfx906</code>, <code class="code">gfx908</code>, <code class="code">gfx90a</code>, <code class="code">gfx90c</code>, 
    691  1.20    mrg <code class="code">gfx1030</code>, <code class="code">gfx1036</code>, <code class="code">gfx1100</code>, <code class="code">gfx1103</code>).
    692  1.19    mrg It ought not include the name of the default
    693  1.20    mrg ISA, specified via <samp class="option">--with-arch</samp>.  If <var class="var">list</var> is empty, then there
    694  1.19    mrg will be no multilibs and only the default run-time library will be built.  If
    695  1.20    mrg <var class="var">list</var> is <code class="code">default</code> or <samp class="option">--with-multilib-list=</samp> is not
    696  1.19    mrg specified, then the default set of libraries is selected.
    697  1.19    mrg </p>
    698  1.19    mrg </dd>
    699  1.20    mrg <dt><code class="code">arm*-*-*</code></dt>
    700  1.20    mrg <dd><p><var class="var">list</var> is a comma separated list of <code class="code">aprofile</code> and
    701  1.20    mrg <code class="code">rmprofile</code> to build multilibs for A or R and M architecture
    702  1.12    mrg profiles respectively.  Note that, due to some limitation of the current
    703  1.20    mrg multilib framework, using the combined <code class="code">aprofile,rmprofile</code>
    704  1.12    mrg multilibs selects in some cases a less optimal multilib than when using
    705  1.12    mrg the multilib profile for the architecture targetted.  The special value
    706  1.20    mrg <code class="code">default</code> is also accepted and is equivalent to omitting the
    707  1.12    mrg option, i.e., only the default run-time library will be enabled.
    708  1.12    mrg </p>
    709  1.20    mrg <p><var class="var">list</var> may instead contain <code class="code">@name</code>, to use the multilib
    710  1.20    mrg configuration Makefile fragment <samp class="file">name</samp> in <samp class="file">gcc/config/arm</samp> in
    711  1.12    mrg the source tree (it is part of the corresponding sources, after all).
    712  1.12    mrg It is recommended, but not required, that files used for this purpose to
    713  1.20    mrg be named starting with <samp class="file">t-ml-</samp>, to make their intended purpose
    714  1.12    mrg self-evident, in line with GCC conventions.  Such files enable custom,
    715  1.12    mrg user-chosen multilib lists to be configured.  Whether multiple such
    716  1.12    mrg files can be used together depends on the contents of the supplied
    717  1.20    mrg files.  See <samp class="file">gcc/config/arm/t-multilib</samp> and its supplementary
    718  1.20    mrg <samp class="file">gcc/config/arm/t-*profile</samp> files for an example of what such
    719  1.12    mrg Makefile fragments might look like for this version of GCC.  The macros
    720  1.12    mrg expected to be defined in these fragments are not stable across GCC
    721  1.20    mrg releases, so make sure they define the <code class="code">MULTILIB</code>-related macros
    722  1.12    mrg expected by the version of GCC you are building.
    723  1.12    mrg See &ldquo;Target Makefile Fragments&rdquo; in the internals manual.
    724  1.10    mrg </p>
    725  1.10    mrg <p>The table below gives the combination of ISAs, architectures, FPUs and
    726  1.12    mrg floating-point ABIs for which multilibs are built for each predefined
    727  1.12    mrg profile.  The union of these options is considered when specifying both
    728  1.20    mrg <code class="code">aprofile</code> and <code class="code">rmprofile</code>.
    729  1.10    mrg </p>
    730  1.20    mrg <table class="multitable">
    731  1.20    mrg <tbody><tr><td width="15%">Option</td><td width="28%">aprofile</td><td width="30%">rmprofile</td></tr>
    732  1.20    mrg <tr><td width="15%">ISAs</td><td width="28%"><code class="code">-marm</code> and <code class="code">-mthumb</code></td><td width="30%"><code class="code">-mthumb</code></td></tr>
    733  1.10    mrg <tr><td width="15%">Architectures<br><br><br><br><br><br></td><td width="28%">default architecture<br>
    734  1.20    mrg <code class="code">-march=armv7-a</code><br>
    735  1.20    mrg <code class="code">-march=armv7ve</code><br>
    736  1.20    mrg <code class="code">-march=armv8-a</code><br><br><br></td><td width="30%">default architecture<br>
    737  1.20    mrg <code class="code">-march=armv6s-m</code><br>
    738  1.20    mrg <code class="code">-march=armv7-m</code><br>
    739  1.20    mrg <code class="code">-march=armv7e-m</code><br>
    740  1.20    mrg <code class="code">-march=armv8-m.base</code><br>
    741  1.20    mrg <code class="code">-march=armv8-m.main</code><br>
    742  1.20    mrg <code class="code">-march=armv7</code></td></tr>
    743  1.10    mrg <tr><td width="15%">FPUs<br><br><br><br><br></td><td width="28%">none<br>
    744  1.20    mrg <code class="code">-mfpu=vfpv3-d16</code><br>
    745  1.20    mrg <code class="code">-mfpu=neon</code><br>
    746  1.20    mrg <code class="code">-mfpu=vfpv4-d16</code><br>
    747  1.20    mrg <code class="code">-mfpu=neon-vfpv4</code><br>
    748  1.20    mrg <code class="code">-mfpu=neon-fp-armv8</code></td><td width="30%">none<br>
    749  1.20    mrg <code class="code">-mfpu=vfpv3-d16</code><br>
    750  1.20    mrg <code class="code">-mfpu=fpv4-sp-d16</code><br>
    751  1.20    mrg <code class="code">-mfpu=fpv5-sp-d16</code><br>
    752  1.20    mrg <code class="code">-mfpu=fpv5-d16</code><br></td></tr>
    753  1.20    mrg <tr><td width="15%">floating-point ABIs<br><br></td><td width="28%"><code class="code">-mfloat-abi=soft</code><br>
    754  1.20    mrg <code class="code">-mfloat-abi=softfp</code><br>
    755  1.20    mrg <code class="code">-mfloat-abi=hard</code></td><td width="30%"><code class="code">-mfloat-abi=soft</code><br>
    756  1.20    mrg <code class="code">-mfloat-abi=softfp</code><br>
    757  1.20    mrg <code class="code">-mfloat-abi=hard</code></td></tr>
    758  1.20    mrg </tbody>
    759  1.10    mrg </table>
    760   1.9    mrg 
    761  1.10    mrg </dd>
    762  1.20    mrg <dt><code class="code">loongarch*-*-*</code></dt>
    763  1.20    mrg <dd><p><var class="var">list</var> is a comma-separated list, with each of the element starting with
    764  1.20    mrg the following ABI identifiers: <code class="code">lp64d[/base]</code> <code class="code">lp64f[/base]</code>
    765  1.20    mrg <code class="code">lp64d[/base]</code> (the <code class="code">/base</code> suffix may be omitted)
    766  1.19    mrg to enable their respective run-time libraries.
    767  1.19    mrg </p>
    768  1.20    mrg <p>A suffix <code class="code">[/<var class="var">arch</var>][/<var class="var">option</var>/&hellip;]</code> may follow immediately
    769  1.19    mrg after the ABI identifier to customize the compiler options for building the
    770  1.20    mrg given set of libraries.  <var class="var">arch</var> denotes the architecture name recognized
    771  1.20    mrg by the <samp class="option">-march=<var class="var">arch</var></samp> compiler option, which acts as a basic target
    772  1.20    mrg ISA configuration that can be adjusted using the subsequent <var class="var">option</var>
    773  1.20    mrg suffixes, where each <var class="var">option</var> is a compiler option without a leading dash
    774  1.19    mrg (&rsquo;-&rsquo;).
    775  1.19    mrg </p>
    776  1.19    mrg <p>If no such suffix is present for a given multilib variant, the
    777  1.20    mrg configured value of <samp class="option">--with-multilib-default</samp> is appended as a default
    778  1.20    mrg suffix.  If <samp class="option">--with-multilib-default</samp> is not given, the default build
    779  1.20    mrg option <samp class="option">-march=abi-default</samp> is applied when building the variants
    780  1.19    mrg without a suffix.
    781  1.19    mrg </p>
    782  1.20    mrg <p>As a special case, <code class="code">fixed</code> may be used in the position of <var class="var">arch</var>,
    783  1.19    mrg which means using the architecture configured with
    784  1.20    mrg <samp class="option">--with-arch=<var class="var">arch</var></samp>, or its default value (e.g. <code class="code">loongarch64</code>
    785  1.20    mrg for <code class="code">loongarch64-*</code> targets).
    786  1.19    mrg </p>
    787  1.20    mrg <p>If <var class="var">list</var> is empty or <code class="code">default</code>, or if <samp class="option">--with-multilib-list</samp>
    788  1.19    mrg is not specified, then only the default variant of the libraries are built,
    789  1.19    mrg where the default ABI is implied by the configured target triplet.
    790  1.16    mrg </p>
    791  1.16    mrg </dd>
    792  1.20    mrg <dt><code class="code">riscv*-*-*</code></dt>
    793  1.20    mrg <dd><p><var class="var">list</var> is a single ABI name.  The target architecture must be either
    794  1.20    mrg <code class="code">rv32gc</code> or <code class="code">rv64gc</code>.  This will build a single multilib for the
    795  1.20    mrg specified architecture and ABI pair.  If <code class="code">--with-multilib-list</code> is not
    796  1.12    mrg given, then a default set of multilibs is selected based on the value of
    797  1.20    mrg <samp class="option">--target</samp>.  This is usually a large set of multilibs.
    798  1.12    mrg </p>
    799  1.12    mrg </dd>
    800  1.20    mrg <dt><code class="code">sh*-*-*</code></dt>
    801  1.20    mrg <dd><p><var class="var">list</var> is a comma separated list of CPU names.  These must be of the
    802  1.20    mrg form <code class="code">sh*</code> or <code class="code">m*</code> (in which case they match the compiler option
    803   1.1    mrg for that processor).  The list should not contain any endian options -
    804  1.20    mrg these are handled by <samp class="option">--with-endian</samp>.
    805  1.10    mrg </p>
    806  1.20    mrg <p>If <var class="var">list</var> is empty, then there will be no multilibs for extra
    807   1.1    mrg processors.  The multilib for the secondary endian remains enabled.
    808  1.10    mrg </p>
    809  1.20    mrg <p>As a special case, if an entry in the list starts with a <code class="code">!</code>
    810  1.10    mrg (exclamation point), then it is added to the list of excluded multilibs.
    811  1.20    mrg Entries of this sort should be compatible with &lsquo;<samp class="samp">MULTILIB_EXCLUDES</samp>&rsquo;
    812  1.20    mrg (once the leading <code class="code">!</code> has been stripped).
    813  1.10    mrg </p>
    814  1.20    mrg <p>If <samp class="option">--with-multilib-list</samp> is not given, then a default set of
    815  1.20    mrg multilibs is selected based on the value of <samp class="option">--target</samp>.  This is
    816   1.1    mrg usually the complete set of libraries, but some targets imply a more
    817   1.1    mrg specialized subset.
    818  1.10    mrg </p>
    819  1.10    mrg <p>Example 1: to configure a compiler for SH4A only, but supporting both
    820  1.10    mrg endians, with little endian being the default:
    821  1.20    mrg </p><div class="example smallexample">
    822  1.20    mrg <pre class="example-preformatted">--with-cpu=sh4a --with-endian=little,big --with-multilib-list=
    823  1.10    mrg </pre></div>
    824   1.9    mrg 
    825  1.10    mrg <p>Example 2: to configure a compiler for both SH4A and SH4AL-DSP, but with
    826   1.1    mrg only little endian SH4AL:
    827  1.20    mrg </p><div class="example smallexample">
    828  1.20    mrg <pre class="example-preformatted">--with-cpu=sh4a --with-endian=little,big \
    829  1.10    mrg --with-multilib-list=sh4al,!mb/m4al
    830  1.10    mrg </pre></div>
    831  1.10    mrg 
    832  1.10    mrg </dd>
    833  1.20    mrg <dt><code class="code">x86-64-*-linux*</code></dt>
    834  1.20    mrg <dd><p><var class="var">list</var> is a comma separated list of <code class="code">m32</code>, <code class="code">m64</code> and
    835  1.20    mrg <code class="code">mx32</code> to enable 32-bit, 64-bit and x32 run-time libraries,
    836  1.20    mrg respectively.  If <var class="var">list</var> is empty, then there will be no multilibs
    837   1.3  skrll and only the default run-time library will be enabled.
    838  1.10    mrg </p>
    839  1.20    mrg <p>If <samp class="option">--with-multilib-list</samp> is not given, then only 32-bit and
    840  1.10    mrg 64-bit run-time libraries will be enabled.
    841  1.10    mrg </p></dd>
    842   1.3  skrll </dl>
    843   1.3  skrll 
    844  1.10    mrg </dd>
    845  1.20    mrg <dt><code class="code">--with-multilib-default</code></dt>
    846  1.19    mrg <dd><p>On LoongArch targets, set the default build options for enabled multilibs
    847  1.19    mrg without build options appended to their corresponding
    848  1.20    mrg <samp class="option">--with-multilib-list</samp> items.  The format of this value is
    849  1.20    mrg <code class="code">[/<var class="var">arch</var>][/<var class="var">option</var>/&hellip;]</code>, where <var class="var">arch</var> is an
    850  1.20    mrg architecture name recognized by <samp class="option">-march=<var class="var">arch</var></samp> compiler option,
    851  1.20    mrg and subsequent <var class="var">option</var> suffixes are compiler options minus a leading
    852  1.19    mrg dash (&rsquo;-&rsquo;).
    853  1.19    mrg </p>
    854  1.20    mrg <p>Multiple <var class="var">option</var>s may appear consecutively while <var class="var">arch</var> may only
    855  1.20    mrg appear in the beginning or be omitted (which means <samp class="option">-march=abi-default</samp>
    856  1.19    mrg is applied when building the libraries).
    857  1.19    mrg </p>
    858  1.19    mrg </dd>
    859  1.20    mrg <dt><code class="code">--with-strict-align-lib</code></dt>
    860  1.20    mrg <dd><p>On LoongArch targets, build all enabled multilibs with <samp class="option">-mstrict-align</samp>
    861  1.19    mrg (Not enabled by default).
    862  1.19    mrg </p>
    863  1.19    mrg </dd>
    864  1.20    mrg <dt><code class="code">--with-multilib-generator=<var class="var">config</var></code></dt>
    865  1.20    mrg <dd><p>Specify what multilibs to build.  <var class="var">config</var> is a semicolon separated list of
    866  1.16    mrg values, possibly consisting of a single value.  Currently only implemented
    867  1.16    mrg for riscv*-*-elf*.  The accepted values and meanings are given below.
    868  1.16    mrg </p>
    869  1.16    mrg 
    870  1.16    mrg <p>Every config is constructed with four components: architecture string, ABI,
    871  1.16    mrg reuse rule with architecture string and reuse rule with sub-extension.
    872  1.16    mrg </p>
    873  1.16    mrg <p>Example 1: Add multi-lib suppport for rv32i with ilp32.
    874  1.20    mrg </p><div class="example smallexample">
    875  1.20    mrg <pre class="example-preformatted">rv32i-ilp32--
    876  1.16    mrg </pre></div>
    877  1.16    mrg 
    878  1.16    mrg <p>Example 2: Add multi-lib suppport for rv32i with ilp32 and rv32imafd with ilp32.
    879  1.20    mrg </p><div class="example smallexample">
    880  1.20    mrg <pre class="example-preformatted">rv32i-ilp32--;rv32imafd-ilp32--
    881  1.16    mrg </pre></div>
    882  1.16    mrg 
    883  1.16    mrg <p>Example 3: Add multi-lib suppport for rv32i with ilp32; rv32im with ilp32 and
    884  1.16    mrg rv32ic with ilp32 will reuse this multi-lib set.
    885  1.20    mrg </p><div class="example smallexample">
    886  1.20    mrg <pre class="example-preformatted">rv32i-ilp32-rv32im-c
    887  1.16    mrg </pre></div>
    888  1.16    mrg 
    889  1.16    mrg <p>Example 4: Add multi-lib suppport for rv64ima with lp64; rv64imaf with lp64,
    890  1.16    mrg rv64imac with lp64 and rv64imafc with lp64 will reuse this multi-lib set.
    891  1.20    mrg </p><div class="example smallexample">
    892  1.20    mrg <pre class="example-preformatted">rv64ima-lp64--f,c,fc
    893  1.16    mrg </pre></div>
    894  1.16    mrg 
    895  1.20    mrg <p><samp class="option">--with-multilib-generator</samp> have an optional configuration argument
    896  1.20    mrg <samp class="option">--cmodel=val</samp> for code model, this option will expand with other
    897  1.20    mrg config options, <var class="var">val</var> is a comma separated list of possible code model,
    898  1.16    mrg currently we support medlow and medany.
    899  1.16    mrg </p>
    900  1.16    mrg <p>Example 5: Add multi-lib suppport for rv64ima with lp64; rv64ima with lp64 and
    901  1.16    mrg medlow code model
    902  1.20    mrg </p><div class="example smallexample">
    903  1.20    mrg <pre class="example-preformatted">rv64ima-lp64--;--cmodel=medlow
    904  1.16    mrg </pre></div>
    905  1.16    mrg 
    906  1.16    mrg <p>Example 6: Add multi-lib suppport for rv64ima with lp64; rv64ima with lp64 and
    907  1.16    mrg medlow code model; rv64ima with lp64 and medany code model
    908  1.20    mrg </p><div class="example smallexample">
    909  1.20    mrg <pre class="example-preformatted">rv64ima-lp64--;--cmodel=medlow,medany
    910  1.16    mrg </pre></div>
    911  1.16    mrg 
    912  1.16    mrg </dd>
    913  1.20    mrg <dt><code class="code">--with-endian=<var class="var">endians</var></code></dt>
    914  1.10    mrg <dd><p>Specify what endians to use.
    915   1.1    mrg Currently only implemented for sh*-*-*.
    916  1.10    mrg </p>
    917  1.20    mrg <p><var class="var">endians</var> may be one of the following:
    918  1.20    mrg </p><dl class="table">
    919  1.20    mrg <dt><code class="code">big</code></dt>
    920  1.10    mrg <dd><p>Use big endian exclusively.
    921  1.10    mrg </p></dd>
    922  1.20    mrg <dt><code class="code">little</code></dt>
    923  1.10    mrg <dd><p>Use little endian exclusively.
    924  1.10    mrg </p></dd>
    925  1.20    mrg <dt><code class="code">big,little</code></dt>
    926  1.10    mrg <dd><p>Use big endian by default.  Provide a multilib for little endian.
    927  1.10    mrg </p></dd>
    928  1.20    mrg <dt><code class="code">little,big</code></dt>
    929  1.10    mrg <dd><p>Use little endian by default.  Provide a multilib for big endian.
    930  1.10    mrg </p></dd>
    931   1.1    mrg </dl>
    932   1.1    mrg 
    933  1.10    mrg </dd>
    934  1.20    mrg <dt><code class="code">--enable-threads</code></dt>
    935  1.10    mrg <dd><p>Specify that the target
    936   1.1    mrg supports threads.  This affects the Objective-C compiler and runtime
    937  1.10    mrg library, and exception handling for other languages like C++.
    938   1.1    mrg On some systems, this is the default.
    939  1.10    mrg </p>
    940  1.10    mrg <p>In general, the best (and, in many cases, the only known) threading
    941   1.1    mrg model available will be configured for use.  Beware that on some
    942   1.1    mrg systems, GCC has not been taught what threading models are generally
    943  1.20    mrg available for the system.  In this case, <samp class="option">--enable-threads</samp> is an
    944  1.20    mrg alias for <samp class="option">--enable-threads=single</samp>.
    945  1.10    mrg </p>
    946  1.10    mrg </dd>
    947  1.20    mrg <dt><code class="code">--disable-threads</code></dt>
    948  1.10    mrg <dd><p>Specify that threading support should be disabled for the system.
    949  1.20    mrg This is an alias for <samp class="option">--enable-threads=single</samp>.
    950  1.10    mrg </p>
    951  1.10    mrg </dd>
    952  1.20    mrg <dt><code class="code">--enable-threads=<var class="var">lib</var></code></dt>
    953  1.10    mrg <dd><p>Specify that
    954  1.20    mrg <var class="var">lib</var> is the thread support library.  This affects the Objective-C
    955   1.1    mrg compiler and runtime library, and exception handling for other languages
    956  1.20    mrg like C++.  The possibilities for <var class="var">lib</var> are:
    957  1.10    mrg </p>
    958  1.20    mrg <dl class="table">
    959  1.20    mrg <dt><code class="code">aix</code></dt>
    960  1.10    mrg <dd><p>AIX thread support.
    961  1.10    mrg </p></dd>
    962  1.20    mrg <dt><code class="code">dce</code></dt>
    963  1.10    mrg <dd><p>DCE thread support.
    964  1.10    mrg </p></dd>
    965  1.20    mrg <dt><code class="code">lynx</code></dt>
    966  1.10    mrg <dd><p>LynxOS thread support.
    967  1.10    mrg </p></dd>
    968  1.20    mrg <dt><code class="code">mipssde</code></dt>
    969  1.10    mrg <dd><p>MIPS SDE thread support.
    970  1.10    mrg </p></dd>
    971  1.20    mrg <dt><code class="code">no</code></dt>
    972  1.20    mrg <dd><p>This is an alias for &lsquo;<samp class="samp">single</samp>&rsquo;.
    973  1.10    mrg </p></dd>
    974  1.20    mrg <dt><code class="code">posix</code></dt>
    975  1.10    mrg <dd><p>Generic POSIX/Unix98 thread support.
    976  1.10    mrg </p></dd>
    977  1.20    mrg <dt><code class="code">rtems</code></dt>
    978  1.10    mrg <dd><p>RTEMS thread support.
    979  1.10    mrg </p></dd>
    980  1.20    mrg <dt><code class="code">single</code></dt>
    981  1.10    mrg <dd><p>Disable thread support, should work for all platforms.
    982  1.10    mrg </p></dd>
    983  1.20    mrg <dt><code class="code">tpf</code></dt>
    984  1.10    mrg <dd><p>TPF thread support.
    985  1.10    mrg </p></dd>
    986  1.20    mrg <dt><code class="code">vxworks</code></dt>
    987  1.10    mrg <dd><p>VxWorks thread support.
    988  1.10    mrg </p></dd>
    989  1.20    mrg <dt><code class="code">win32</code></dt>
    990  1.10    mrg <dd><p>Microsoft Win32 API thread support.
    991  1.10    mrg </p></dd>
    992   1.1    mrg </dl>
    993   1.1    mrg 
    994  1.10    mrg </dd>
    995  1.20    mrg <dt><code class="code">--enable-tls</code></dt>
    996  1.10    mrg <dd><p>Specify that the target supports TLS (Thread Local Storage).  Usually
    997   1.1    mrg configure can correctly determine if TLS is supported.  In cases where
    998   1.1    mrg it guesses incorrectly, TLS can be explicitly enabled or disabled with
    999  1.20    mrg <samp class="option">--enable-tls</samp> or <samp class="option">--disable-tls</samp>.  This can happen if
   1000   1.1    mrg the assembler supports TLS but the C library does not, or if the
   1001   1.1    mrg assumptions made by the configure test are incorrect.
   1002  1.10    mrg </p>
   1003  1.10    mrg </dd>
   1004  1.20    mrg <dt><code class="code">--disable-tls</code></dt>
   1005  1.10    mrg <dd><p>Specify that the target does not support TLS.
   1006  1.20    mrg This is an alias for <samp class="option">--enable-tls=no</samp>.
   1007  1.10    mrg </p>
   1008  1.10    mrg </dd>
   1009  1.20    mrg <dt><code class="code">--disable-tm-clone-registry</code></dt>
   1010  1.13    mrg <dd><p>Disable TM clone registry in libgcc. It is enabled in libgcc by default.
   1011  1.13    mrg This option helps to reduce code size for embedded targets which do
   1012  1.13    mrg not use transactional memory.
   1013  1.13    mrg </p>
   1014  1.13    mrg </dd>
   1015  1.20    mrg <dt><code class="code">--with-cpu=<var class="var">cpu</var></code></dt>
   1016  1.20    mrg <dt><code class="code">--with-cpu-32=<var class="var">cpu</var></code></dt>
   1017  1.20    mrg <dt><code class="code">--with-cpu-64=<var class="var">cpu</var></code></dt>
   1018  1.10    mrg <dd><p>Specify which cpu variant the compiler should generate code for by default.
   1019  1.20    mrg <var class="var">cpu</var> will be used as the default value of the <samp class="option">-mcpu=</samp> switch.
   1020   1.5    mrg This option is only supported on some targets, including ARC, ARM, i386, M68k,
   1021  1.20    mrg PowerPC, and SPARC.  It is mandatory for ARC.  The <samp class="option">--with-cpu-32</samp> and
   1022  1.20    mrg <samp class="option">--with-cpu-64</samp> options specify separate default CPUs for
   1023  1.13    mrg 32-bit and 64-bit modes; these options are only supported for aarch64, i386,
   1024   1.7    mrg x86-64, PowerPC, and SPARC.
   1025  1.10    mrg </p>
   1026  1.10    mrg </dd>
   1027  1.20    mrg <dt><code class="code">--with-schedule=<var class="var">cpu</var></code></dt>
   1028  1.20    mrg <dt><code class="code">--with-arch=<var class="var">cpu</var></code></dt>
   1029  1.20    mrg <dt><code class="code">--with-arch-32=<var class="var">cpu</var></code></dt>
   1030  1.20    mrg <dt><code class="code">--with-arch-64=<var class="var">cpu</var></code></dt>
   1031  1.20    mrg <dt><code class="code">--with-tune=<var class="var">cpu</var></code></dt>
   1032  1.20    mrg <dt><code class="code">--with-tune-32=<var class="var">cpu</var></code></dt>
   1033  1.20    mrg <dt><code class="code">--with-tune-64=<var class="var">cpu</var></code></dt>
   1034  1.20    mrg <dt><code class="code">--with-abi=<var class="var">abi</var></code></dt>
   1035  1.20    mrg <dt><code class="code">--with-fpu=<var class="var">type</var></code></dt>
   1036  1.20    mrg <dt><code class="code">--with-float=<var class="var">type</var></code></dt>
   1037  1.20    mrg <dt><code class="code">--with-simd=<var class="var">type</var></code></dt>
   1038  1.20    mrg <dd><p>These configure options provide default values for the <samp class="option">-mschedule=</samp>,
   1039  1.20    mrg <samp class="option">-march=</samp>, <samp class="option">-mtune=</samp>, <samp class="option">-mabi=</samp>, and <samp class="option">-mfpu=</samp>
   1040  1.20    mrg options and for <samp class="option">-mhard-float</samp> or <samp class="option">-msoft-float</samp>.  As with
   1041  1.20    mrg <samp class="option">--with-cpu</samp>, which switches will be accepted and acceptable values
   1042   1.1    mrg of the arguments depend on the target.
   1043  1.10    mrg </p>
   1044  1.10    mrg </dd>
   1045  1.20    mrg <dt><code class="code">--with-mode=<var class="var">mode</var></code></dt>
   1046  1.20    mrg <dd><p>Specify if the compiler should default to <samp class="option">-marm</samp> or <samp class="option">-mthumb</samp>.
   1047   1.1    mrg This option is only supported on ARM targets.
   1048  1.10    mrg </p>
   1049  1.10    mrg </dd>
   1050  1.20    mrg <dt><code class="code">--with-stack-offset=<var class="var">num</var></code></dt>
   1051  1.20    mrg <dd><p>This option sets the default for the -mstack-offset=<var class="var">num</var> option,
   1052   1.3  skrll and will thus generally also control the setting of this option for
   1053   1.3  skrll libraries.  This option is only supported on Epiphany targets.
   1054  1.10    mrg </p>
   1055  1.10    mrg </dd>
   1056  1.20    mrg <dt><code class="code">--with-fpmath=<var class="var">isa</var></code></dt>
   1057  1.20    mrg <dd><p>This options sets <samp class="option">-mfpmath=sse</samp> by default and specifies the default
   1058  1.20    mrg ISA for floating-point arithmetics.  You can select either &lsquo;<samp class="samp">sse</samp>&rsquo; which
   1059  1.20    mrg enables <samp class="option">-msse2</samp> or &lsquo;<samp class="samp">avx</samp>&rsquo; which enables <samp class="option">-mavx</samp> by default.
   1060   1.3  skrll This option is only supported on i386 and x86-64 targets.
   1061  1.10    mrg </p>
   1062  1.10    mrg </dd>
   1063  1.20    mrg <dt><code class="code">--with-fp-32=<var class="var">mode</var></code></dt>
   1064  1.20    mrg <dd><p>On MIPS targets, set the default value for the <samp class="option">-mfp</samp> option when using
   1065  1.20    mrg the o32 ABI.  The possibilities for <var class="var">mode</var> are:
   1066  1.20    mrg </p><dl class="table">
   1067  1.20    mrg <dt><code class="code">32</code></dt>
   1068  1.20    mrg <dd><p>Use the o32 FP32 ABI extension, as with the <samp class="option">-mfp32</samp> command-line
   1069  1.10    mrg option.
   1070  1.10    mrg </p></dd>
   1071  1.20    mrg <dt><code class="code">xx</code></dt>
   1072  1.20    mrg <dd><p>Use the o32 FPXX ABI extension, as with the <samp class="option">-mfpxx</samp> command-line
   1073  1.10    mrg option.
   1074  1.10    mrg </p></dd>
   1075  1.20    mrg <dt><code class="code">64</code></dt>
   1076  1.20    mrg <dd><p>Use the o32 FP64 ABI extension, as with the <samp class="option">-mfp64</samp> command-line
   1077  1.10    mrg option.
   1078  1.10    mrg </p></dd>
   1079   1.5    mrg </dl>
   1080  1.10    mrg <p>In the absence of this configuration option the default is to use the o32
   1081   1.5    mrg FP32 ABI extension.
   1082  1.10    mrg </p>
   1083  1.10    mrg </dd>
   1084  1.20    mrg <dt><code class="code">--with-odd-spreg-32</code></dt>
   1085  1.20    mrg <dd><p>On MIPS targets, set the <samp class="option">-modd-spreg</samp> option by default when using
   1086   1.5    mrg the o32 ABI.
   1087  1.10    mrg </p>
   1088  1.10    mrg </dd>
   1089  1.20    mrg <dt><code class="code">--without-odd-spreg-32</code></dt>
   1090  1.20    mrg <dd><p>On MIPS targets, set the <samp class="option">-mno-odd-spreg</samp> option by default when using
   1091   1.5    mrg the o32 ABI.  This is normally used in conjunction with
   1092  1.20    mrg <samp class="option">--with-fp-32=64</samp> in order to target the o32 FP64A ABI extension.
   1093  1.10    mrg </p>
   1094  1.10    mrg </dd>
   1095  1.20    mrg <dt><code class="code">--with-nan=<var class="var">encoding</var></code></dt>
   1096  1.10    mrg <dd><p>On MIPS targets, set the default encoding convention to use for the
   1097   1.5    mrg special not-a-number (NaN) IEEE 754 floating-point data.  The
   1098  1.20    mrg possibilities for <var class="var">encoding</var> are:
   1099  1.20    mrg </p><dl class="table">
   1100  1.20    mrg <dt><code class="code">legacy</code></dt>
   1101  1.20    mrg <dd><p>Use the legacy encoding, as with the <samp class="option">-mnan=legacy</samp> command-line
   1102  1.10    mrg option.
   1103  1.10    mrg </p></dd>
   1104  1.20    mrg <dt><code class="code">2008</code></dt>
   1105  1.20    mrg <dd><p>Use the 754-2008 encoding, as with the <samp class="option">-mnan=2008</samp> command-line
   1106  1.10    mrg option.
   1107  1.10    mrg </p></dd>
   1108   1.5    mrg </dl>
   1109  1.10    mrg <p>To use this configuration option you must have an assembler version
   1110  1.20    mrg installed that supports the <samp class="option">-mnan=</samp> command-line option too.
   1111   1.5    mrg In the absence of this configuration option the default convention is
   1112  1.20    mrg the legacy encoding, as when neither of the <samp class="option">-mnan=2008</samp> and
   1113  1.20    mrg <samp class="option">-mnan=legacy</samp> command-line options has been used.
   1114  1.10    mrg </p>
   1115  1.10    mrg </dd>
   1116  1.20    mrg <dt><code class="code">--with-divide=<var class="var">type</var></code></dt>
   1117  1.10    mrg <dd><p>Specify how the compiler should generate code for checking for
   1118  1.10    mrg division by zero.  This option is only supported on the MIPS target.
   1119  1.20    mrg The possibilities for <var class="var">type</var> are:
   1120  1.20    mrg </p><dl class="table">
   1121  1.20    mrg <dt><code class="code">traps</code></dt>
   1122  1.10    mrg <dd><p>Division by zero checks use conditional traps (this is the default on
   1123  1.10    mrg systems that support conditional traps).
   1124  1.10    mrg </p></dd>
   1125  1.20    mrg <dt><code class="code">breaks</code></dt>
   1126  1.10    mrg <dd><p>Division by zero checks use the break instruction.
   1127  1.10    mrg </p></dd>
   1128   1.1    mrg </dl>
   1129   1.1    mrg 
   1130  1.19    mrg </dd>
   1131  1.20    mrg <dt><code class="code">--with-compact-branches=<var class="var">policy</var></code></dt>
   1132  1.19    mrg <dd><p>Specify how the compiler should generate branch instructions.
   1133  1.19    mrg This option is only supported on the MIPS target.
   1134  1.20    mrg The possibilities for <var class="var">type</var> are:
   1135  1.20    mrg </p><dl class="table">
   1136  1.20    mrg <dt><code class="code">optimal</code></dt>
   1137  1.19    mrg <dd><p>Cause a delay slot branch to be used if one is available in the
   1138  1.19    mrg current ISA and the delay slot is successfully filled. If the delay slot
   1139  1.19    mrg is not filled, a compact branch will be chosen if one is available.
   1140  1.19    mrg </p></dd>
   1141  1.20    mrg <dt><code class="code">never</code></dt>
   1142  1.19    mrg <dd><p>Ensures that compact branch instructions will never be generated.
   1143  1.19    mrg </p></dd>
   1144  1.20    mrg <dt><code class="code">always</code></dt>
   1145  1.19    mrg <dd><p>Ensures that a compact branch instruction will be generated if available.
   1146  1.19    mrg If a compact branch instruction is not available,
   1147  1.19    mrg a delay slot form of the branch will be used instead.
   1148  1.19    mrg This option is supported from MIPS Release 6 onwards.
   1149  1.19    mrg For pre-R6/microMIPS/MIPS16, this option is just same as never/optimal.
   1150  1.19    mrg </p></dd>
   1151  1.19    mrg </dl>
   1152  1.19    mrg 
   1153  1.10    mrg 
   1154  1.10    mrg </dd>
   1155  1.20    mrg <dt><code class="code">--with-llsc</code></dt>
   1156  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-mllsc</samp> the default when no
   1157  1.20    mrg <samp class="option">-mno-llsc</samp> option is passed.  This is the default for
   1158   1.1    mrg Linux-based targets, as the kernel will emulate them if the ISA does
   1159   1.1    mrg not provide them.
   1160  1.10    mrg </p>
   1161  1.10    mrg </dd>
   1162  1.20    mrg <dt><code class="code">--without-llsc</code></dt>
   1163  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-mno-llsc</samp> the default when no
   1164  1.20    mrg <samp class="option">-mllsc</samp> option is passed.
   1165  1.10    mrg </p>
   1166  1.10    mrg </dd>
   1167  1.20    mrg <dt><code class="code">--with-synci</code></dt>
   1168  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-msynci</samp> the default when no
   1169  1.20    mrg <samp class="option">-mno-synci</samp> option is passed.
   1170  1.10    mrg </p>
   1171  1.10    mrg </dd>
   1172  1.20    mrg <dt><code class="code">--without-synci</code></dt>
   1173  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-mno-synci</samp> the default when no
   1174  1.20    mrg <samp class="option">-msynci</samp> option is passed.  This is the default.
   1175  1.10    mrg </p>
   1176  1.10    mrg </dd>
   1177  1.20    mrg <dt><code class="code">--with-lxc1-sxc1</code></dt>
   1178  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-mlxc1-sxc1</samp> the default when no
   1179  1.20    mrg <samp class="option">-mno-lxc1-sxc1</samp> option is passed.  This is the default.
   1180  1.10    mrg </p>
   1181  1.10    mrg </dd>
   1182  1.20    mrg <dt><code class="code">--without-lxc1-sxc1</code></dt>
   1183  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-mno-lxc1-sxc1</samp> the default when no
   1184  1.20    mrg <samp class="option">-mlxc1-sxc1</samp> option is passed.  The indexed load/store
   1185   1.9    mrg instructions are not directly a problem but can lead to unexpected
   1186   1.9    mrg behaviour when deployed in an application intended for a 32-bit address
   1187   1.9    mrg space but run on a 64-bit processor.  The issue is seen because all
   1188   1.9    mrg known MIPS 64-bit Linux kernels execute o32 and n32 applications
   1189   1.9    mrg with 64-bit addressing enabled which affects the overflow behaviour
   1190   1.9    mrg of the indexed addressing mode.  GCC will assume that ordinary
   1191   1.9    mrg 32-bit arithmetic overflow behaviour is the same whether performed
   1192  1.20    mrg as an <code class="code">addu</code> instruction or as part of the address calculation
   1193  1.20    mrg in <code class="code">lwxc1</code> type instructions.  This assumption holds true in a
   1194   1.9    mrg pure 32-bit environment and can hold true in a 64-bit environment if
   1195   1.9    mrg the address space is accurately set to be 32-bit for o32 and n32.
   1196  1.10    mrg </p>
   1197  1.10    mrg </dd>
   1198  1.20    mrg <dt><code class="code">--with-madd4</code></dt>
   1199  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-mmadd4</samp> the default when no
   1200  1.20    mrg <samp class="option">-mno-madd4</samp> option is passed.  This is the default.
   1201  1.10    mrg </p>
   1202  1.10    mrg </dd>
   1203  1.20    mrg <dt><code class="code">--without-madd4</code></dt>
   1204  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-mno-madd4</samp> the default when no
   1205  1.20    mrg <samp class="option">-mmadd4</samp> option is passed.  The <code class="code">madd4</code> instruction
   1206   1.9    mrg family can be problematic when targeting a combination of cores that
   1207   1.9    mrg implement these instructions differently.  There are two known cores
   1208   1.9    mrg that implement these as fused operations instead of unfused (where
   1209   1.9    mrg unfused is normally expected).  Disabling these instructions is the
   1210   1.9    mrg only way to ensure compatible code is generated; this will incur
   1211   1.9    mrg a performance penalty.
   1212  1.10    mrg </p>
   1213  1.10    mrg </dd>
   1214  1.20    mrg <dt><code class="code">--with-msa</code></dt>
   1215  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-mmsa</samp> the default when no
   1216  1.20    mrg <samp class="option">-mno-msa</samp> option is passed.
   1217  1.19    mrg </p>
   1218  1.19    mrg </dd>
   1219  1.20    mrg <dt><code class="code">--without-msa</code></dt>
   1220  1.20    mrg <dd><p>On MIPS targets, make <samp class="option">-mno-msa</samp> the default when no
   1221  1.20    mrg <samp class="option">-mmsa</samp> option is passed. This is the default.
   1222  1.19    mrg </p>
   1223  1.19    mrg </dd>
   1224  1.20    mrg <dt><code class="code">--with-mips-plt</code></dt>
   1225  1.10    mrg <dd><p>On MIPS targets, make use of copy relocations and PLTs.
   1226   1.1    mrg These features are extensions to the traditional
   1227   1.1    mrg SVR4-based MIPS ABIs and require support from GNU binutils
   1228   1.1    mrg and the runtime C library.
   1229  1.10    mrg </p>
   1230  1.10    mrg </dd>
   1231  1.20    mrg <dt><code class="code">--with-stack-clash-protection-guard-size=<var class="var">size</var></code></dt>
   1232  1.12    mrg <dd><p>On certain targets this option sets the default stack clash protection guard
   1233  1.20    mrg size as a power of two in bytes.  On AArch64 <var class="var">size</var> is required to be either
   1234  1.12    mrg 12 (4KB) or 16 (64KB).
   1235  1.12    mrg </p>
   1236  1.12    mrg </dd>
   1237  1.20    mrg <dt><code class="code">--with-isa-spec=<var class="var">ISA-spec-string</var></code></dt>
   1238  1.16    mrg <dd><p>On RISC-V targets specify the default version of the RISC-V Unprivileged
   1239  1.16    mrg (formerly User-Level) ISA specification to produce code conforming to.
   1240  1.20    mrg The possibilities for <var class="var">ISA-spec-string</var> are:
   1241  1.20    mrg </p><dl class="table">
   1242  1.20    mrg <dt><code class="code">2.2</code></dt>
   1243  1.16    mrg <dd><p>Produce code conforming to version 2.2.
   1244  1.16    mrg </p></dd>
   1245  1.20    mrg <dt><code class="code">20190608</code></dt>
   1246  1.16    mrg <dd><p>Produce code conforming to version 20190608.
   1247  1.16    mrg </p></dd>
   1248  1.20    mrg <dt><code class="code">20191213</code></dt>
   1249  1.16    mrg <dd><p>Produce code conforming to version 20191213.
   1250  1.16    mrg </p></dd>
   1251  1.16    mrg </dl>
   1252  1.16    mrg <p>In the absence of this configuration option the default version is 20191213.
   1253  1.16    mrg </p>
   1254  1.16    mrg </dd>
   1255  1.20    mrg <dt><code class="code">--enable-__cxa_atexit</code></dt>
   1256  1.10    mrg <dd><p>Define if you want to use __cxa_atexit, rather than atexit, to
   1257  1.10    mrg register C++ destructors for local statics and global objects.
   1258   1.1    mrg This is essential for fully standards-compliant handling of
   1259   1.1    mrg destructors, but requires __cxa_atexit in libc.  This option is currently
   1260   1.1    mrg only available on systems with GNU libc.  When enabled, this will cause
   1261  1.20    mrg <samp class="option">-fuse-cxa-atexit</samp> to be passed by default.
   1262  1.10    mrg </p>
   1263  1.10    mrg </dd>
   1264  1.20    mrg <dt><code class="code">--enable-gnu-indirect-function</code></dt>
   1265  1.20    mrg <dd><p>Define if you want to enable the <code class="code">ifunc</code> attribute.  This option is
   1266   1.3  skrll currently only available on systems with GNU libc on certain targets.
   1267  1.10    mrg </p>
   1268  1.10    mrg </dd>
   1269  1.20    mrg <dt><code class="code">--enable-target-optspace</code></dt>
   1270  1.10    mrg <dd><p>Specify that target
   1271  1.10    mrg libraries should be optimized for code space instead of code speed.
   1272   1.1    mrg This is the default for the m32r platform.
   1273  1.10    mrg </p>
   1274  1.10    mrg </dd>
   1275  1.20    mrg <dt><code class="code">--with-cpp-install-dir=<var class="var">dirname</var></code></dt>
   1276  1.20    mrg <dd><p>Specify that the user visible <code class="command">cpp</code> program should be installed
   1277  1.20    mrg in <samp class="file"><var class="var">prefix</var>/<var class="var">dirname</var>/cpp</samp>, in addition to <var class="var">bindir</var>.
   1278  1.10    mrg </p>
   1279  1.10    mrg </dd>
   1280  1.20    mrg <dt><code class="code">--enable-comdat</code></dt>
   1281  1.10    mrg <dd><p>Enable COMDAT group support.  This is primarily used to override the
   1282   1.1    mrg automatically detected value.
   1283  1.10    mrg </p>
   1284  1.10    mrg </dd>
   1285  1.20    mrg <dt><code class="code">--enable-initfini-array</code></dt>
   1286  1.20    mrg <dd><p>Force the use of sections <code class="code">.init_array</code> and <code class="code">.fini_array</code>
   1287  1.20    mrg (instead of <code class="code">.init</code> and <code class="code">.fini</code>) for constructors and
   1288  1.20    mrg destructors.  Option <samp class="option">--disable-initfini-array</samp> has the
   1289   1.1    mrg opposite effect.  If neither option is specified, the configure script
   1290  1.20    mrg will try to guess whether the <code class="code">.init_array</code> and
   1291  1.20    mrg <code class="code">.fini_array</code> sections are supported and, if they are, use them.
   1292  1.10    mrg </p>
   1293  1.10    mrg </dd>
   1294  1.20    mrg <dt><code class="code">--enable-link-mutex</code></dt>
   1295  1.10    mrg <dd><p>When building GCC, use a mutex to avoid linking the compilers for
   1296   1.5    mrg multiple languages at the same time, to avoid thrashing on build
   1297   1.5    mrg systems with limited free memory.  The default is not to use such a mutex.
   1298  1.10    mrg </p>
   1299  1.10    mrg </dd>
   1300  1.20    mrg <dt><code class="code">--enable-link-serialization</code></dt>
   1301  1.16    mrg <dd><p>When building GCC, use make dependencies to serialize linking the compilers for
   1302  1.16    mrg multiple languages, to avoid thrashing on build
   1303  1.16    mrg systems with limited free memory.  The default is not to add such
   1304  1.16    mrg dependencies and thus with parallel make potentially link different
   1305  1.16    mrg compilers concurrently.  If the argument is a positive integer, allow
   1306  1.16    mrg that number of concurrent link processes for the large binaries.
   1307  1.16    mrg </p>
   1308  1.16    mrg </dd>
   1309  1.20    mrg <dt><code class="code">--enable-maintainer-mode</code></dt>
   1310  1.10    mrg <dd><p>The build rules that regenerate the Autoconf and Automake output files as
   1311  1.20    mrg well as the GCC master message catalog <samp class="file">gcc.pot</samp> are normally
   1312   1.1    mrg disabled.  This is because it can only be rebuilt if the complete source
   1313   1.1    mrg tree is present.  If you have changed the sources and want to rebuild the
   1314  1.20    mrg catalog, configuring with <samp class="option">--enable-maintainer-mode</samp> will enable
   1315  1.20    mrg this.  Note that you need a recent version of the <code class="code">gettext</code> tools
   1316   1.1    mrg to do so.
   1317  1.10    mrg </p>
   1318  1.10    mrg </dd>
   1319  1.20    mrg <dt><code class="code">--disable-bootstrap</code></dt>
   1320  1.10    mrg <dd><p>For a native build, the default configuration is to perform
   1321  1.20    mrg a 3-stage bootstrap of the compiler when &lsquo;<samp class="samp">make</samp>&rsquo; is invoked,
   1322   1.1    mrg testing that GCC can compile itself correctly.  If you want to disable
   1323  1.20    mrg this process, you can configure with <samp class="option">--disable-bootstrap</samp>.
   1324  1.10    mrg </p>
   1325  1.10    mrg </dd>
   1326  1.20    mrg <dt><code class="code">--enable-bootstrap</code></dt>
   1327  1.10    mrg <dd><p>In special cases, you may want to perform a 3-stage build
   1328  1.10    mrg even if the target and host triplets are different.
   1329   1.1    mrg This is possible when the host can run code compiled for
   1330  1.10    mrg the target (e.g. host is i686-linux, target is i486-linux).
   1331   1.1    mrg Starting from GCC 4.2, to do this you have to configure explicitly
   1332  1.20    mrg with <samp class="option">--enable-bootstrap</samp>.
   1333  1.10    mrg </p>
   1334  1.10    mrg </dd>
   1335  1.20    mrg <dt><code class="code">--enable-generated-files-in-srcdir</code></dt>
   1336  1.10    mrg <dd><p>Neither the .c and .h files that are generated from Bison and flex nor the
   1337   1.1    mrg info manuals and man pages that are built from the .texi files are present
   1338  1.11    mrg in the repository development tree.  When building GCC from that development tree,
   1339   1.1    mrg or from one of our snapshots, those generated files are placed in your
   1340   1.1    mrg build directory, which allows for the source to be in a readonly
   1341   1.1    mrg directory.
   1342  1.10    mrg </p>
   1343  1.20    mrg <p>If you configure with <samp class="option">--enable-generated-files-in-srcdir</samp> then those
   1344   1.1    mrg generated files will go into the source directory.  This is mainly intended
   1345   1.1    mrg for generating release or prerelease tarballs of the GCC sources, since it
   1346   1.1    mrg is not a requirement that the users of source releases to have flex, Bison,
   1347   1.1    mrg or makeinfo.
   1348  1.10    mrg </p>
   1349  1.10    mrg </dd>
   1350  1.20    mrg <dt><code class="code">--enable-version-specific-runtime-libs</code></dt>
   1351  1.10    mrg <dd><p>Specify
   1352   1.1    mrg that runtime libraries should be installed in the compiler specific
   1353  1.20    mrg subdirectory (<samp class="file"><var class="var">libdir</var>/gcc</samp>) rather than the usual places.  In
   1354  1.20    mrg addition, &lsquo;<samp class="samp">libstdc++</samp>&rsquo;&rsquo;s include files will be installed into
   1355  1.20    mrg <samp class="file"><var class="var">libdir</var></samp> unless you overruled it by using
   1356  1.20    mrg <samp class="option">--with-gxx-include-dir=<var class="var">dirname</var></samp>.  Using this option is
   1357   1.1    mrg particularly useful if you intend to use several versions of GCC in
   1358  1.20    mrg parallel.  The default is &lsquo;<samp class="samp">yes</samp>&rsquo; for &lsquo;<samp class="samp">libada</samp>&rsquo;, and &lsquo;<samp class="samp">no</samp>&rsquo; for
   1359  1.13    mrg the remaining libraries.
   1360  1.10    mrg </p>
   1361  1.10    mrg </dd>
   1362  1.20    mrg <dt><code class="code">--with-darwin-extra-rpath</code></dt>
   1363  1.19    mrg <dd><p>This is provided to allow distributions to add a single additional
   1364  1.19    mrg runpath on Darwin / macOS systems. This allows for cases where the
   1365  1.19    mrg installed GCC library directories are then symlinked to a common
   1366  1.19    mrg directory outside of the GCC installation.
   1367  1.19    mrg </p>
   1368  1.19    mrg </dd>
   1369  1.20    mrg <dt><code class="code"><a class="anchor" id="WithAixSoname"></a>--with-aix-soname=&lsquo;<samp class="samp">aix</samp>&rsquo;, &lsquo;<samp class="samp">svr4</samp>&rsquo; or &lsquo;<samp class="samp">both</samp>&rsquo;</code></dt>
   1370  1.20    mrg <dd><p>Traditional AIX shared library versioning (versioned <code class="code">Shared Object</code>
   1371  1.20    mrg files as members of unversioned <code class="code">Archive Library</code> files named
   1372  1.20    mrg &lsquo;<samp class="samp">lib.a</samp>&rsquo;) causes numerous headaches for package managers. However,
   1373  1.20    mrg <code class="code">Import Files</code> as members of <code class="code">Archive Library</code> files allow for
   1374  1.20    mrg <strong class="strong">filename-based versioning</strong> of shared libraries as seen on Linux/SVR4,
   1375  1.10    mrg where this is called the &quot;SONAME&quot;. But as they prevent static linking,
   1376  1.20    mrg <code class="code">Import Files</code> may be used with <code class="code">Runtime Linking</code> only, where the
   1377  1.20    mrg linker does search for &lsquo;<samp class="samp">libNAME.so</samp>&rsquo; before &lsquo;<samp class="samp">libNAME.a</samp>&rsquo; library
   1378  1.20    mrg filenames with the &lsquo;<samp class="samp">-lNAME</samp>&rsquo; linker flag.
   1379  1.10    mrg </p>
   1380  1.20    mrg <a class="anchor" id="AixLdCommand"></a><p>For detailed information please refer to the AIX
   1381  1.20    mrg <a class="uref" href="https://www.ibm.com/support/knowledgecenter/search/%22the%20ld%20command%2C%20also%20called%20the%20linkage%20editor%20or%20binder%22">ld
   1382  1.10    mrg Command</a> reference.
   1383  1.10    mrg </p>
   1384  1.10    mrg <p>As long as shared library creation is enabled, upon:
   1385  1.20    mrg </p><dl class="table">
   1386  1.20    mrg <dt><code class="code">--with-aix-soname=aix</code></dt>
   1387  1.20    mrg <dt><code class="code">--with-aix-soname=both</code></dt>
   1388  1.20    mrg <dd><p>A (traditional AIX) <code class="code">Shared Archive Library</code> file is created:
   1389  1.20    mrg  </p><ul class="itemize mark-bullet">
   1390  1.20    mrg <li>using the &lsquo;<samp class="samp">libNAME.a</samp>&rsquo; filename scheme
   1391  1.20    mrg   </li><li>with the <code class="code">Shared Object</code> file as archive member named
   1392  1.20    mrg   &lsquo;<samp class="samp">libNAME.so.V</samp>&rsquo; (except for &lsquo;<samp class="samp">libgcc_s</samp>&rsquo;, where the <code class="code">Shared
   1393  1.20    mrg   Object</code> file is named &lsquo;<samp class="samp">shr.o</samp>&rsquo; for backwards compatibility), which
   1394  1.20    mrg   <ul class="itemize mark-minus">
   1395  1.20    mrg <li>is used for runtime loading from inside the &lsquo;<samp class="samp">libNAME.a</samp>&rsquo; file
   1396  1.20    mrg    </li><li>is used for dynamic loading via
   1397  1.20    mrg    <code class="code">dlopen(&quot;libNAME.a(libNAME.so.V)&quot;, RTLD_MEMBER)</code>
   1398  1.20    mrg    </li><li>is used for shared linking
   1399  1.20    mrg    </li><li>is used for static linking, so no separate <code class="code">Static Archive
   1400   1.5    mrg    Library</code> file is needed
   1401  1.10    mrg   </li></ul>
   1402  1.10    mrg </li></ul>
   1403  1.10    mrg </dd>
   1404  1.20    mrg <dt><code class="code">--with-aix-soname=both</code></dt>
   1405  1.20    mrg <dt><code class="code">--with-aix-soname=svr4</code></dt>
   1406  1.20    mrg <dd><p>A (second) <code class="code">Shared Archive Library</code> file is created:
   1407  1.20    mrg  </p><ul class="itemize mark-bullet">
   1408  1.20    mrg <li>using the &lsquo;<samp class="samp">libNAME.so.V</samp>&rsquo; filename scheme
   1409  1.20    mrg  </li><li>with the <code class="code">Shared Object</code> file as archive member named
   1410  1.20    mrg  &lsquo;<samp class="samp">shr.o</samp>&rsquo;, which
   1411  1.20    mrg   <ul class="itemize mark-minus">
   1412  1.20    mrg <li>is created with the <code class="code">-G linker flag</code>
   1413  1.20    mrg    </li><li>has the <code class="code">F_LOADONLY</code> flag set
   1414  1.20    mrg    </li><li>is used for runtime loading from inside the &lsquo;<samp class="samp">libNAME.so.V</samp>&rsquo; file
   1415  1.20    mrg    </li><li>is used for dynamic loading via <code class="code">dlopen(&quot;libNAME.so.V(shr.o)&quot;,
   1416   1.5    mrg    RTLD_MEMBER)</code>
   1417  1.10    mrg   </li></ul>
   1418  1.20    mrg </li><li>with the <code class="code">Import File</code> as archive member named &lsquo;<samp class="samp">shr.imp</samp>&rsquo;,
   1419   1.5    mrg  which
   1420  1.20    mrg   <ul class="itemize mark-minus">
   1421  1.20    mrg <li>refers to &lsquo;<samp class="samp">libNAME.so.V(shr.o)</samp>&rsquo; as the &quot;SONAME&quot;, to be recorded
   1422  1.20    mrg    in the <code class="code">Loader Section</code> of subsequent binaries
   1423  1.20    mrg    </li><li>indicates whether &lsquo;<samp class="samp">libNAME.so.V(shr.o)</samp>&rsquo; is 32 or 64 bit
   1424  1.20    mrg    </li><li>lists all the public symbols exported by &lsquo;<samp class="samp">lib.so.V(shr.o)</samp>&rsquo;,
   1425  1.20    mrg    eventually decorated with the <code class="code">&lsquo;<samp class="samp">weak</samp>&rsquo; Keyword</code>
   1426  1.20    mrg    </li><li>is necessary for shared linking against &lsquo;<samp class="samp">lib.so.V(shr.o)</samp>&rsquo;
   1427  1.10    mrg    </li></ul>
   1428  1.10    mrg </li></ul>
   1429  1.20    mrg <p>A symbolic link using the &lsquo;<samp class="samp">libNAME.so</samp>&rsquo; filename scheme is created:
   1430  1.20    mrg   </p><ul class="itemize mark-bullet">
   1431  1.20    mrg <li>pointing to the &lsquo;<samp class="samp">libNAME.so.V</samp>&rsquo; <code class="code">Shared Archive Library</code> file
   1432  1.20    mrg   </li><li>to permit the <code class="code">ld Command</code> to find &lsquo;<samp class="samp">lib.so.V(shr.imp)</samp>&rsquo; via
   1433  1.20    mrg   the &lsquo;<samp class="samp">-lNAME</samp>&rsquo; argument (requires <code class="code">Runtime Linking</code> to be enabled)
   1434  1.20    mrg   </li><li>to permit dynamic loading of &lsquo;<samp class="samp">lib.so.V(shr.o)</samp>&rsquo; without the need
   1435  1.20    mrg   to specify the version number via <code class="code">dlopen(&quot;libNAME.so(shr.o)&quot;,
   1436   1.5    mrg   RTLD_MEMBER)</code>
   1437  1.10    mrg   </li></ul>
   1438  1.10    mrg </dd>
   1439  1.10    mrg </dl>
   1440   1.5    mrg 
   1441  1.10    mrg <p>As long as static library creation is enabled, upon:
   1442  1.20    mrg </p><dl class="table">
   1443  1.20    mrg <dt><code class="code">--with-aix-soname=svr4</code></dt>
   1444  1.20    mrg <dd><p>A <code class="code">Static Archive Library</code> is created:
   1445  1.20    mrg  </p><ul class="itemize mark-bullet">
   1446  1.20    mrg <li>using the &lsquo;<samp class="samp">libNAME.a</samp>&rsquo; filename scheme
   1447  1.20    mrg  </li><li>with all the <code class="code">Static Object</code> files as archive members, which
   1448  1.20    mrg   <ul class="itemize mark-minus">
   1449  1.20    mrg <li>are used for static linking
   1450  1.10    mrg   </li></ul>
   1451  1.10    mrg </li></ul>
   1452  1.10    mrg </dd>
   1453  1.10    mrg </dl>
   1454   1.5    mrg 
   1455  1.20    mrg <p>While the aix-soname=&lsquo;<samp class="samp">svr4</samp>&rsquo; option does not create <code class="code">Shared Object</code>
   1456  1.20    mrg files as members of unversioned <code class="code">Archive Library</code> files any more, package
   1457   1.5    mrg managers still are responsible to
   1458  1.20    mrg <a class="uref" href="./specific.html#TransferAixShobj">transfer</a> <code class="code">Shared Object</code> files
   1459  1.20    mrg found as member of a previously installed unversioned <code class="code">Archive Library</code>
   1460  1.20    mrg file into the newly installed <code class="code">Archive Library</code> file with the same
   1461   1.5    mrg filename.
   1462  1.10    mrg </p>
   1463  1.20    mrg <p><em class="emph">WARNING:</em> Creating <code class="code">Shared Object</code> files with <code class="code">Runtime Linking</code>
   1464  1.20    mrg enabled may bloat the TOC, eventually leading to <code class="code">TOC overflow</code> errors,
   1465  1.20    mrg requiring the use of either the <samp class="option">-Wl,-bbigtoc</samp> linker flag (seen to
   1466  1.20    mrg break with the <code class="code">GDB</code> debugger) or some of the TOC-related compiler flags,
   1467   1.5    mrg see &ldquo;RS/6000 and PowerPC Options&rdquo; in the main manual.
   1468  1.10    mrg </p>
   1469  1.20    mrg <p><samp class="option">--with-aix-soname</samp> is currently supported by &lsquo;<samp class="samp">libgcc_s</samp>&rsquo; only, so
   1470   1.5    mrg this option is still experimental and not for normal use yet.
   1471  1.10    mrg </p>
   1472  1.20    mrg <p>Default is the traditional behavior <samp class="option">--with-aix-soname=&lsquo;<samp class="samp">aix</samp>&rsquo;</samp>.
   1473  1.10    mrg </p>
   1474  1.10    mrg </dd>
   1475  1.20    mrg <dt><code class="code">--enable-languages=<var class="var">lang1</var>,<var class="var">lang2</var>,&hellip;</code></dt>
   1476  1.10    mrg <dd><p>Specify that only a particular subset of compilers and
   1477   1.1    mrg their runtime libraries should be built.  For a list of valid values for
   1478  1.20    mrg <var class="var">langN</var> you can issue the following command in the
   1479  1.20    mrg <samp class="file">gcc</samp> directory of your GCC source tree:<br>
   1480  1.20    mrg </p><div class="example smallexample">
   1481  1.20    mrg <pre class="example-preformatted">grep ^language= */config-lang.in
   1482  1.10    mrg </pre></div>
   1483  1.10    mrg <p>Currently, you can use any of the following:
   1484  1.20    mrg <code class="code">all</code>, <code class="code">default</code>, <code class="code">ada</code>, <code class="code">c</code>, <code class="code">c++</code>, <code class="code">d</code>,
   1485  1.20    mrg <code class="code">fortran</code>, <code class="code">go</code>, <code class="code">jit</code>, <code class="code">lto</code>, <code class="code">m2</code>,
   1486  1.20    mrg <code class="code">objc</code>, <code class="code">obj-c++</code>.
   1487  1.10    mrg Building the Ada compiler has special requirements, see below.
   1488  1.20    mrg If you do not pass this flag, or specify the option <code class="code">default</code>, then the
   1489  1.20    mrg default languages available in the <samp class="file">gcc</samp> sub-tree will be configured.
   1490  1.19    mrg Ada, D, Go, Jit, Objective-C++ and Modula-2 are not default languages.
   1491  1.19    mrg LTO is not a
   1492  1.20    mrg default language, but is built by default because <samp class="option">--enable-lto</samp> is
   1493  1.10    mrg enabled by default.  The other languages are default languages.  If
   1494  1.20    mrg <code class="code">all</code> is specified, then all available languages are built.  An
   1495  1.20    mrg exception is <code class="code">jit</code> language, which requires
   1496  1.20    mrg <samp class="option">--enable-host-shared</samp> to be included with <code class="code">all</code>.
   1497  1.10    mrg </p>
   1498  1.10    mrg </dd>
   1499  1.20    mrg <dt><code class="code">--enable-stage1-languages=<var class="var">lang1</var>,<var class="var">lang2</var>,&hellip;</code></dt>
   1500  1.10    mrg <dd><p>Specify that a particular subset of compilers and their runtime
   1501   1.1    mrg libraries should be built with the system C compiler during stage 1 of
   1502   1.1    mrg the bootstrap process, rather than only in later stages with the
   1503   1.1    mrg bootstrapped C compiler.  The list of valid values is the same as for
   1504  1.20    mrg <samp class="option">--enable-languages</samp>, and the option <code class="code">all</code> will select all
   1505  1.20    mrg of the languages enabled by <samp class="option">--enable-languages</samp>.  This option is
   1506   1.1    mrg primarily useful for GCC development; for instance, when a development
   1507   1.1    mrg version of the compiler cannot bootstrap due to compiler bugs, or when
   1508   1.1    mrg one is debugging front ends other than the C front end.  When this
   1509   1.1    mrg option is used, one can then build the target libraries for the
   1510  1.20    mrg specified languages with the stage-1 compiler by using <code class="command">make
   1511  1.10    mrg stage1-bubble all-target</code>, or run the testsuite on the stage-1 compiler
   1512  1.20    mrg for the specified languages using <code class="command">make stage1-start check-gcc</code>.
   1513  1.10    mrg </p>
   1514  1.10    mrg </dd>
   1515  1.20    mrg <dt><code class="code">--disable-libada</code></dt>
   1516  1.10    mrg <dd><p>Specify that the run-time libraries and tools used by GNAT should not
   1517   1.1    mrg be built.  This can be useful for debugging, or for compatibility with
   1518   1.1    mrg previous Ada build procedures, when it was required to explicitly
   1519  1.20    mrg do a &lsquo;<samp class="samp">make -C gcc gnatlib_and_tools</samp>&rsquo;.
   1520  1.10    mrg </p>
   1521  1.10    mrg </dd>
   1522  1.20    mrg <dt><code class="code">--disable-libgm2</code></dt>
   1523  1.19    mrg <dd><p>Specify that the run-time libraries and tools used by Modula-2 should not
   1524  1.19    mrg be built.  This can be useful for debugging.
   1525  1.19    mrg </p>
   1526  1.19    mrg </dd>
   1527  1.20    mrg <dt><code class="code">--disable-libsanitizer</code></dt>
   1528  1.10    mrg <dd><p>Specify that the run-time libraries for the various sanitizers should
   1529   1.5    mrg not be built.
   1530  1.10    mrg </p>
   1531  1.10    mrg </dd>
   1532  1.20    mrg <dt><code class="code">--disable-libssp</code></dt>
   1533  1.10    mrg <dd><p>Specify that the run-time libraries for stack smashing protection
   1534  1.10    mrg should not be built or linked against.  On many targets library support
   1535  1.10    mrg is provided by the C library instead.
   1536  1.10    mrg </p>
   1537  1.10    mrg </dd>
   1538  1.20    mrg <dt><code class="code">--disable-libquadmath</code></dt>
   1539  1.10    mrg <dd><p>Specify that the GCC quad-precision math library should not be built.
   1540   1.3  skrll On some systems, the library is required to be linkable when building
   1541  1.20    mrg the Fortran front end, unless <samp class="option">--disable-libquadmath-support</samp>
   1542   1.3  skrll is used.
   1543  1.10    mrg </p>
   1544  1.10    mrg </dd>
   1545  1.20    mrg <dt><code class="code">--disable-libquadmath-support</code></dt>
   1546  1.20    mrg <dd><p>Specify that the Fortran front end and <code class="code">libgfortran</code> do not add
   1547  1.20    mrg support for <code class="code">libquadmath</code> on systems supporting it.
   1548  1.10    mrg </p>
   1549  1.10    mrg </dd>
   1550  1.20    mrg <dt><code class="code">--disable-libgomp</code></dt>
   1551  1.10    mrg <dd><p>Specify that the GNU Offloading and Multi Processing Runtime Library
   1552   1.5    mrg should not be built.
   1553  1.10    mrg </p>
   1554  1.10    mrg </dd>
   1555  1.20    mrg <dt><code class="code">--disable-libvtv</code></dt>
   1556  1.10    mrg <dd><p>Specify that the run-time libraries used by vtable verification
   1557   1.5    mrg should not be built.
   1558  1.10    mrg </p>
   1559  1.10    mrg </dd>
   1560  1.20    mrg <dt><code class="code">--with-dwarf2</code></dt>
   1561  1.10    mrg <dd><p>Specify that the compiler should
   1562  1.19    mrg use DWARF debugging information as the default; the exact
   1563  1.19    mrg DWARF version that is the default is target-specific.
   1564  1.10    mrg </p>
   1565  1.10    mrg </dd>
   1566  1.20    mrg <dt><code class="code">--with-advance-toolchain=<var class="var">at</var></code></dt>
   1567  1.10    mrg <dd><p>On 64-bit PowerPC Linux systems, configure the compiler to use the
   1568   1.7    mrg header files, library files, and the dynamic linker from the Advance
   1569  1.20    mrg Toolchain release <var class="var">at</var> instead of the default versions that are
   1570   1.7    mrg provided by the Linux distribution.  In general, this option is
   1571   1.7    mrg intended for the developers of GCC, and it is not intended for general
   1572   1.7    mrg use.
   1573  1.10    mrg </p>
   1574  1.10    mrg </dd>
   1575  1.20    mrg <dt><code class="code">--enable-targets=all</code></dt>
   1576  1.20    mrg <dt><code class="code">--enable-targets=<var class="var">target_list</var></code></dt>
   1577  1.10    mrg <dd><p>Some GCC targets, e.g. powerpc64-linux, build bi-arch compilers.
   1578   1.1    mrg These are compilers that are able to generate either 64-bit or 32-bit
   1579  1.10    mrg code.  Typically, the corresponding 32-bit target, e.g.
   1580   1.1    mrg powerpc-linux for powerpc64-linux, only generates 32-bit code.  This
   1581   1.1    mrg option enables the 32-bit target to be a bi-arch compiler, which is
   1582   1.1    mrg useful when you want a bi-arch compiler that defaults to 32-bit, and
   1583  1.10    mrg you are building a bi-arch or multi-arch binutils in a combined tree.
   1584   1.1    mrg On mips-linux, this will build a tri-arch compiler (ABI o32/n32/64),
   1585  1.10    mrg defaulted to o32.
   1586   1.3  skrll Currently, this option only affects sparc-linux, powerpc-linux, x86-linux,
   1587   1.3  skrll mips-linux and s390-linux.
   1588  1.10    mrg </p>
   1589  1.10    mrg </dd>
   1590  1.20    mrg <dt><code class="code">--enable-default-pie</code></dt>
   1591  1.20    mrg <dd><p>Turn on <samp class="option">-fPIE</samp> and <samp class="option">-pie</samp> by default.
   1592  1.10    mrg </p>
   1593  1.10    mrg </dd>
   1594  1.20    mrg <dt><code class="code">--enable-secureplt</code></dt>
   1595  1.20    mrg <dd><p>This option enables <samp class="option">-msecure-plt</samp> by default for powerpc-linux.
   1596   1.1    mrg See &ldquo;RS/6000 and PowerPC Options&rdquo; in the main manual
   1597  1.10    mrg </p>
   1598  1.10    mrg </dd>
   1599  1.20    mrg <dt><code class="code">--enable-default-ssp</code></dt>
   1600  1.20    mrg <dd><p>Turn on <samp class="option">-fstack-protector-strong</samp> by default.
   1601  1.10    mrg </p>
   1602  1.10    mrg </dd>
   1603  1.20    mrg <dt><code class="code">--enable-cld</code></dt>
   1604  1.20    mrg <dd><p>This option enables <samp class="option">-mcld</samp> by default for 32-bit x86 targets.
   1605   1.1    mrg See &ldquo;i386 and x86-64 Options&rdquo; in the main manual
   1606  1.10    mrg </p>
   1607  1.10    mrg </dd>
   1608  1.20    mrg <dt><code class="code">--enable-large-address-aware</code></dt>
   1609  1.20    mrg <dd><p>The <samp class="option">--enable-large-address-aware</samp> option arranges for MinGW
   1610  1.20    mrg executables to be linked using the <samp class="option">--large-address-aware</samp>
   1611  1.12    mrg option, that enables the use of more than 2GB of memory.  If GCC is
   1612  1.12    mrg configured with this option, its effects can be reversed by passing the
   1613  1.20    mrg <samp class="option">-Wl,--disable-large-address-aware</samp> option to the so-configured
   1614  1.12    mrg compiler driver.
   1615  1.12    mrg </p>
   1616  1.12    mrg </dd>
   1617  1.20    mrg <dt><code class="code">--enable-win32-registry</code></dt>
   1618  1.20    mrg <dt><code class="code">--enable-win32-registry=<var class="var">key</var></code></dt>
   1619  1.20    mrg <dt><code class="code">--disable-win32-registry</code></dt>
   1620  1.20    mrg <dd><p>The <samp class="option">--enable-win32-registry</samp> option enables Microsoft Windows-hosted GCC
   1621   1.1    mrg to look up installations paths in the registry using the following key:
   1622  1.10    mrg </p>
   1623  1.20    mrg <div class="example smallexample">
   1624  1.20    mrg <pre class="example-preformatted"><code class="code">HKEY_LOCAL_MACHINE\SOFTWARE\Free Software Foundation\<var class="var">key</var></code>
   1625  1.10    mrg </pre></div>
   1626   1.1    mrg 
   1627  1.20    mrg <p><var class="var">key</var> defaults to GCC version number, and can be overridden by the
   1628  1.20    mrg <samp class="option">--enable-win32-registry=<var class="var">key</var></samp> option.  Vendors and distributors
   1629   1.1    mrg who use custom installers are encouraged to provide a different key,
   1630   1.1    mrg perhaps one comprised of vendor name and GCC version number, to
   1631   1.1    mrg avoid conflict with existing installations.  This feature is enabled
   1632  1.20    mrg by default, and can be disabled by <samp class="option">--disable-win32-registry</samp>
   1633   1.1    mrg option.  This option has no effect on the other hosts.
   1634  1.10    mrg </p>
   1635  1.10    mrg </dd>
   1636  1.20    mrg <dt><code class="code">--nfp</code></dt>
   1637  1.10    mrg <dd><p>Specify that the machine does not have a floating point unit.  This
   1638  1.20    mrg option only applies to &lsquo;<samp class="samp">m68k-sun-sunos<var class="var">n</var></samp>&rsquo;.  On any other
   1639  1.20    mrg system, <samp class="option">--nfp</samp> has no effect.
   1640  1.10    mrg </p>
   1641  1.10    mrg </dd>
   1642  1.20    mrg <dt><code class="code">--enable-werror</code></dt>
   1643  1.20    mrg <dt><code class="code">--disable-werror</code></dt>
   1644  1.20    mrg <dt><code class="code">--enable-werror=yes</code></dt>
   1645  1.20    mrg <dt><code class="code">--enable-werror=no</code></dt>
   1646  1.10    mrg <dd><p>When you specify this option, it controls whether certain files in the
   1647  1.20    mrg compiler are built with <samp class="option">-Werror</samp> in bootstrap stage2 and later.
   1648  1.20    mrg If you don&rsquo;t specify it, <samp class="option">-Werror</samp> is turned on for the main
   1649   1.1    mrg development trunk.  However it defaults to off for release branches and
   1650  1.20    mrg final releases.  The specific files which get <samp class="option">-Werror</samp> are
   1651   1.1    mrg controlled by the Makefiles.
   1652  1.10    mrg </p>
   1653  1.10    mrg </dd>
   1654  1.20    mrg <dt><code class="code">--enable-checking</code></dt>
   1655  1.20    mrg <dt><code class="code">--disable-checking</code></dt>
   1656  1.20    mrg <dt><code class="code">--enable-checking=<var class="var">list</var></code></dt>
   1657  1.11    mrg <dd><p>This option controls performing internal consistency checks in the compiler.
   1658  1.11    mrg It does not change the generated code, but adds error checking of the
   1659  1.11    mrg requested complexity.  This slows down the compiler and may only work
   1660  1.11    mrg properly if you are building the compiler with GCC.
   1661  1.11    mrg </p>
   1662  1.11    mrg <p>When the option is not specified, the active set of checks depends on context.
   1663  1.20    mrg Namely, bootstrap stage 1 defaults to &lsquo;<samp class="samp">--enable-checking=yes</samp>&rsquo;, builds
   1664  1.11    mrg from release branches or release archives default to
   1665  1.20    mrg &lsquo;<samp class="samp">--enable-checking=release</samp>&rsquo;, and otherwise
   1666  1.20    mrg &lsquo;<samp class="samp">--enable-checking=yes,extra</samp>&rsquo; is used.  When the option is
   1667  1.20    mrg specified without a <var class="var">list</var>, the result is the same as
   1668  1.20    mrg &lsquo;<samp class="samp">--enable-checking=yes</samp>&rsquo;.  Likewise, &lsquo;<samp class="samp">--disable-checking</samp>&rsquo; is
   1669  1.20    mrg equivalent to &lsquo;<samp class="samp">--enable-checking=no</samp>&rsquo;.
   1670  1.20    mrg </p>
   1671  1.20    mrg <p>The categories of checks available in <var class="var">list</var> are &lsquo;<samp class="samp">yes</samp>&rsquo; (most common
   1672  1.20    mrg checks &lsquo;<samp class="samp">assert,misc,gc,gimple,rtlflag,runtime,tree,types</samp>&rsquo;), &lsquo;<samp class="samp">no</samp>&rsquo;
   1673  1.20    mrg (no checks at all), &lsquo;<samp class="samp">all</samp>&rsquo; (all but &lsquo;<samp class="samp">valgrind</samp>&rsquo;), &lsquo;<samp class="samp">release</samp>&rsquo;
   1674  1.20    mrg (cheapest checks &lsquo;<samp class="samp">assert,runtime</samp>&rsquo;) or &lsquo;<samp class="samp">none</samp>&rsquo; (same as &lsquo;<samp class="samp">no</samp>&rsquo;).
   1675  1.20    mrg &lsquo;<samp class="samp">release</samp>&rsquo; checks are always on and to disable them
   1676  1.20    mrg &lsquo;<samp class="samp">--disable-checking</samp>&rsquo; or &lsquo;<samp class="samp">--enable-checking=no[,&lt;other checks&gt;]</samp>&rsquo;
   1677  1.11    mrg must be explicitly requested.  Disabling assertions makes the compiler and
   1678  1.11    mrg runtime slightly faster but increases the risk of undetected internal errors
   1679  1.11    mrg causing wrong code to be generated.
   1680  1.11    mrg </p>
   1681  1.20    mrg <p>Individual checks can be enabled with these flags: &lsquo;<samp class="samp">assert</samp>&rsquo;, &lsquo;<samp class="samp">df</samp>&rsquo;,
   1682  1.20    mrg &lsquo;<samp class="samp">extra</samp>&rsquo;, &lsquo;<samp class="samp">fold</samp>&rsquo;, &lsquo;<samp class="samp">gc</samp>&rsquo;, &lsquo;<samp class="samp">gcac</samp>&rsquo;, &lsquo;<samp class="samp">gimple</samp>&rsquo;,
   1683  1.20    mrg &lsquo;<samp class="samp">misc</samp>&rsquo;, &lsquo;<samp class="samp">rtl</samp>&rsquo;, &lsquo;<samp class="samp">rtlflag</samp>&rsquo;, &lsquo;<samp class="samp">runtime</samp>&rsquo;, &lsquo;<samp class="samp">tree</samp>&rsquo;,
   1684  1.20    mrg &lsquo;<samp class="samp">types</samp>&rsquo; and &lsquo;<samp class="samp">valgrind</samp>&rsquo;.  &lsquo;<samp class="samp">extra</samp>&rsquo; extends &lsquo;<samp class="samp">misc</samp>&rsquo;
   1685  1.11    mrg checking with extra checks that might affect code generation and should
   1686  1.11    mrg therefore not differ between stage1 and later stages in bootstrap.
   1687  1.11    mrg </p>
   1688  1.20    mrg <p>The &lsquo;<samp class="samp">valgrind</samp>&rsquo; check requires the external <code class="command">valgrind</code> simulator,
   1689  1.20    mrg available from <a class="uref" href="https://valgrind.org">https://valgrind.org</a>.  The &lsquo;<samp class="samp">rtl</samp>&rsquo; checks are
   1690  1.20    mrg expensive and the &lsquo;<samp class="samp">df</samp>&rsquo;, &lsquo;<samp class="samp">gcac</samp>&rsquo; and &lsquo;<samp class="samp">valgrind</samp>&rsquo; checks are very
   1691  1.11    mrg expensive.
   1692  1.10    mrg </p>
   1693  1.10    mrg </dd>
   1694  1.20    mrg <dt><code class="code">--disable-stage1-checking</code></dt>
   1695  1.20    mrg <dt><code class="code">--enable-stage1-checking</code></dt>
   1696  1.20    mrg <dt><code class="code">--enable-stage1-checking=<var class="var">list</var></code></dt>
   1697  1.20    mrg <dd><p>This option affects only bootstrap build.  If no <samp class="option">--enable-checking</samp>
   1698  1.20    mrg option is specified the stage1 compiler is built with &lsquo;<samp class="samp">yes</samp>&rsquo; checking
   1699  1.11    mrg enabled, otherwise the stage1 checking flags are the same as specified by
   1700  1.20    mrg <samp class="option">--enable-checking</samp>.  To build the stage1 compiler with
   1701  1.20    mrg different checking options use <samp class="option">--enable-stage1-checking</samp>.
   1702  1.20    mrg The list of checking options is the same as for <samp class="option">--enable-checking</samp>.
   1703   1.1    mrg If your system is too slow or too small to bootstrap a released compiler
   1704  1.20    mrg with checking for stage1 enabled, you can use &lsquo;<samp class="samp">--disable-stage1-checking</samp>&rsquo;
   1705   1.1    mrg to disable checking for the stage1 compiler.
   1706  1.10    mrg </p>
   1707  1.10    mrg </dd>
   1708  1.20    mrg <dt><code class="code">--enable-coverage</code></dt>
   1709  1.20    mrg <dt><code class="code">--enable-coverage=<var class="var">level</var></code></dt>
   1710  1.10    mrg <dd><p>With this option, the compiler is built to collect self coverage
   1711   1.1    mrg information, every time it is run.  This is for internal development
   1712   1.1    mrg purposes, and only works when the compiler is being built with gcc.  The
   1713  1.20    mrg <var class="var">level</var> argument controls whether the compiler is built optimized or
   1714  1.20    mrg not, values are &lsquo;<samp class="samp">opt</samp>&rsquo; and &lsquo;<samp class="samp">noopt</samp>&rsquo;.  For coverage analysis you
   1715   1.1    mrg want to disable optimization, for performance analysis you want to
   1716   1.1    mrg enable optimization.  When coverage is enabled, the default level is
   1717   1.1    mrg without optimization.
   1718  1.10    mrg </p>
   1719  1.10    mrg </dd>
   1720  1.20    mrg <dt><code class="code">--enable-gather-detailed-mem-stats</code></dt>
   1721  1.10    mrg <dd><p>When this option is specified more detailed information on memory
   1722   1.1    mrg allocation is gathered.  This information is printed when using
   1723  1.20    mrg <samp class="option">-fmem-report</samp>.
   1724  1.10    mrg </p>
   1725  1.10    mrg </dd>
   1726  1.20    mrg <dt><code class="code">--enable-valgrind-annotations</code></dt>
   1727  1.10    mrg <dd><p>Mark selected memory related operations in the compiler when run under
   1728   1.7    mrg valgrind to suppress false positives.
   1729  1.10    mrg </p>
   1730  1.10    mrg </dd>
   1731  1.20    mrg <dt><code class="code">--enable-nls</code></dt>
   1732  1.20    mrg <dt><code class="code">--disable-nls</code></dt>
   1733  1.20    mrg <dd><p>The <samp class="option">--enable-nls</samp> option enables Native Language Support (NLS),
   1734   1.1    mrg which lets GCC output diagnostics in languages other than American
   1735   1.1    mrg English.  Native Language Support is enabled by default if not doing a
   1736  1.20    mrg canadian cross build.  The <samp class="option">--disable-nls</samp> option disables NLS.
   1737  1.10    mrg </p>
   1738  1.19    mrg <p>Note that this functionality requires either libintl (provided by GNU
   1739  1.19    mrg gettext) or C standard library that contains support for gettext (such
   1740  1.19    mrg as the GNU C Library).
   1741  1.20    mrg See <a class="xref" href="#with-included-gettext">&ndash;with-included-gettext</a> for more
   1742  1.19    mrg information on the conditions required to get gettext support.
   1743  1.19    mrg </p>
   1744  1.19    mrg </dd>
   1745  1.20    mrg <dt><code class="code">--with-libintl-prefix=<var class="var">dir</var></code></dt>
   1746  1.20    mrg <dt><code class="code">--without-libintl-prefix</code></dt>
   1747  1.20    mrg <dd><p>Searches for libintl in <samp class="file"><var class="var">dir</var>/include</samp> and
   1748  1.20    mrg <samp class="file"><var class="var">dir</var>/lib</samp>, or disables manual searching for it, letting the
   1749  1.19    mrg linker handle it.
   1750  1.19    mrg </p>
   1751  1.10    mrg </dd>
   1752  1.20    mrg <dt><code class="code">--with-libintl-type=<var class="var">type</var></code></dt>
   1753  1.19    mrg <dd><p>Specifies the type of library to search for when looking for libintl.
   1754  1.20    mrg <var class="var">type</var> can be one of <code class="code">auto</code>, <code class="code">static</code> or <code class="code">shared</code>.
   1755  1.19    mrg </p>
   1756  1.20    mrg <a class="anchor" id="with-included-gettext"></a></dd>
   1757  1.20    mrg <dt><code class="code">--with-included-gettext</code></dt>
   1758  1.20    mrg <dd><p>Only available if <samp class="file">gettext</samp> is present in the source tree.
   1759  1.10    mrg </p>
   1760  1.19    mrg <p>Forces the gettext tree to be configured to build and use a new static
   1761  1.19    mrg libintl, overriding the system libintl.  Results in GCC being built
   1762  1.19    mrg against the newly built libintl rather than the system libintl.
   1763  1.19    mrg </p>
   1764  1.19    mrg <p>The build system makes a somewhat complicated choice when picking where
   1765  1.19    mrg to get gettext routines from.  The following table is a summary of the
   1766  1.19    mrg possible options:
   1767  1.19    mrg </p>
   1768  1.20    mrg <table class="multitable">
   1769  1.20    mrg <thead><tr><th width="12%">GNU gettext present in sources</th><th width="12%">libintl installed on the system</th><th width="12%"><code class="code">gettext</code> present in libc</th><th width="20%"><code class="code">--with-included-gettext</code></th><th width="44%">Effects on localization</th></tr></thead>
   1770  1.20    mrg <tbody><tr><td width="12%">No</td><td width="12%">No</td><td width="12%">No</td><td width="20%">(ignored)</td><td width="44%">No localization</td></tr>
   1771  1.19    mrg <tr><td width="12%">No</td><td width="12%">No</td><td width="12%">Yes</td><td width="20%">(ignored)</td><td width="44%">Localized, libc gettext</td></tr>
   1772  1.19    mrg <tr><td width="12%">No</td><td width="12%">Yes</td><td width="12%">No</td><td width="20%">(ignored)</td><td width="44%">Localized, libintl</td></tr>
   1773  1.19    mrg <tr><td width="12%">No</td><td width="12%">Yes</td><td width="12%">Yes</td><td width="20%">(ignored)</td><td width="44%">Localized, libintl</td></tr>
   1774  1.19    mrg <tr><td width="12%">Yes</td><td width="12%">No</td><td width="12%">No</td><td width="20%">No</td><td width="44%">Localized, new, static libintl</td></tr>
   1775  1.19    mrg <tr><td width="12%">Yes</td><td width="12%">No</td><td width="12%">No</td><td width="20%">Yes</td><td width="44%">Localized, new, static libintl</td></tr>
   1776  1.19    mrg <tr><td width="12%">Yes</td><td width="12%">No</td><td width="12%">Yes</td><td width="20%">No</td><td width="44%">Localized, libc gettext</td></tr>
   1777  1.19    mrg <tr><td width="12%">Yes</td><td width="12%">No</td><td width="12%">Yes</td><td width="20%">Yes</td><td width="44%">Localized, new, static libintl</td></tr>
   1778  1.19    mrg <tr><td width="12%">Yes</td><td width="12%">Yes</td><td width="12%">No</td><td width="20%">No</td><td width="44%">Localized, libintl</td></tr>
   1779  1.19    mrg <tr><td width="12%">Yes</td><td width="12%">Yes</td><td width="12%">No</td><td width="20%">Yes</td><td width="44%">Localized, new, static libintl</td></tr>
   1780  1.19    mrg <tr><td width="12%">Yes</td><td width="12%">Yes</td><td width="12%">Yes</td><td width="20%">No</td><td width="44%">Localized, libintl</td></tr>
   1781  1.19    mrg <tr><td width="12%">Yes</td><td width="12%">Yes</td><td width="12%">Yes</td><td width="20%">Yes</td><td width="44%">Localized, new, static libintl</td></tr>
   1782  1.20    mrg </tbody>
   1783  1.19    mrg </table>
   1784  1.19    mrg 
   1785  1.10    mrg </dd>
   1786  1.20    mrg <dt><code class="code">--with-catgets</code></dt>
   1787  1.20    mrg <dd><p>If NLS is enabled, and if the host lacks <code class="code">gettext</code> but has the
   1788  1.20    mrg inferior <code class="code">catgets</code> interface, the GCC build procedure normally
   1789  1.20    mrg ignores <code class="code">catgets</code> and instead uses GCC&rsquo;s copy of the GNU
   1790  1.20    mrg <code class="code">gettext</code> library.  The <samp class="option">--with-catgets</samp> option causes the
   1791  1.20    mrg build procedure to use the host&rsquo;s <code class="code">catgets</code> in this situation.
   1792  1.10    mrg </p>
   1793  1.10    mrg </dd>
   1794  1.20    mrg <dt><code class="code">--with-libiconv-prefix=<var class="var">dir</var></code></dt>
   1795  1.20    mrg <dd><p>Search for libiconv header files in <samp class="file"><var class="var">dir</var>/include</samp> and
   1796  1.20    mrg libiconv library files in <samp class="file"><var class="var">dir</var>/lib</samp>.
   1797  1.10    mrg </p>
   1798  1.10    mrg </dd>
   1799  1.20    mrg <dt><code class="code">--enable-obsolete</code></dt>
   1800  1.10    mrg <dd><p>Enable configuration for an obsoleted system.  If you attempt to
   1801   1.1    mrg configure GCC for a system (build, host, or target) which has been
   1802   1.1    mrg obsoleted, and you do not specify this flag, configure will halt with an
   1803   1.1    mrg error message.
   1804  1.10    mrg </p>
   1805  1.10    mrg <p>All support for systems which have been obsoleted in one release of GCC
   1806   1.1    mrg is removed entirely in the next major release, unless someone steps
   1807   1.1    mrg forward to maintain the port.
   1808  1.10    mrg </p>
   1809  1.10    mrg </dd>
   1810  1.20    mrg <dt><code class="code">--enable-decimal-float</code></dt>
   1811  1.20    mrg <dt><code class="code">--enable-decimal-float=yes</code></dt>
   1812  1.20    mrg <dt><code class="code">--enable-decimal-float=no</code></dt>
   1813  1.20    mrg <dt><code class="code">--enable-decimal-float=bid</code></dt>
   1814  1.20    mrg <dt><code class="code">--enable-decimal-float=dpd</code></dt>
   1815  1.20    mrg <dt><code class="code">--disable-decimal-float</code></dt>
   1816  1.10    mrg <dd><p>Enable (or disable) support for the C decimal floating point extension
   1817  1.19    mrg that is in the IEEE 754-2008 standard.  This is enabled by default
   1818  1.19    mrg only on AArch64, PowerPC, i386, and x86_64 GNU/Linux systems.  Other
   1819  1.19    mrg systems may also support it, but require the user to specifically
   1820  1.19    mrg enable it.  You can optionally control which decimal floating point
   1821  1.20    mrg format is used (either &lsquo;<samp class="samp">bid</samp>&rsquo; or &lsquo;<samp class="samp">dpd</samp>&rsquo;).  The &lsquo;<samp class="samp">bid</samp>&rsquo;
   1822  1.19    mrg (binary integer decimal) format is default on AArch64, i386 and x86_64
   1823  1.20    mrg systems, and the &lsquo;<samp class="samp">dpd</samp>&rsquo; (densely packed decimal) format is default
   1824  1.19    mrg on PowerPC systems.
   1825  1.10    mrg </p>
   1826  1.10    mrg </dd>
   1827  1.20    mrg <dt><code class="code">--enable-fixed-point</code></dt>
   1828  1.20    mrg <dt><code class="code">--disable-fixed-point</code></dt>
   1829  1.10    mrg <dd><p>Enable (or disable) support for C fixed-point arithmetic.
   1830   1.1    mrg This option is enabled by default for some targets (such as MIPS) which
   1831   1.1    mrg have hardware-support for fixed-point operations.  On other targets, you
   1832   1.1    mrg may enable this option manually.
   1833  1.10    mrg </p>
   1834  1.10    mrg </dd>
   1835  1.20    mrg <dt><code class="code">--with-long-double-128</code></dt>
   1836  1.20    mrg <dd><p>Specify if <code class="code">long double</code> type should be 128-bit by default on selected
   1837  1.20    mrg GNU/Linux architectures.  If using <code class="code">--without-long-double-128</code>,
   1838  1.20    mrg <code class="code">long double</code> will be by default 64-bit, the same as <code class="code">double</code> type.
   1839   1.1    mrg When neither of these configure options are used, the default will be
   1840  1.20    mrg 128-bit <code class="code">long double</code> when built against GNU C Library 2.4 and later,
   1841  1.20    mrg 64-bit <code class="code">long double</code> otherwise.
   1842  1.10    mrg </p>
   1843  1.10    mrg </dd>
   1844  1.20    mrg <dt><code class="code">--with-long-double-format=ibm</code></dt>
   1845  1.20    mrg <dt><code class="code">--with-long-double-format=ieee</code></dt>
   1846  1.20    mrg <dd><p>Specify whether <code class="code">long double</code> uses the IBM extended double format
   1847  1.10    mrg or the IEEE 128-bit floating point format on PowerPC Linux systems.
   1848  1.10    mrg This configuration switch will only work on little endian PowerPC
   1849  1.10    mrg Linux systems and on big endian 64-bit systems where the default cpu
   1850  1.20    mrg is at least power7 (i.e. <samp class="option">--with-cpu=power7</samp>,
   1851  1.20    mrg <samp class="option">--with-cpu=power8</samp>, or <samp class="option">--with-cpu=power9</samp> is used).
   1852  1.10    mrg </p>
   1853  1.20    mrg <p>If you use the <samp class="option">--with-long-double-64</samp> configuration option,
   1854  1.20    mrg the <samp class="option">--with-long-double-format=ibm</samp> and
   1855  1.20    mrg <samp class="option">--with-long-double-format=ieee</samp> options are ignored.
   1856  1.10    mrg </p>
   1857  1.20    mrg <p>The default <code class="code">long double</code> format is to use IBM extended double.
   1858  1.10    mrg Until all of the libraries are converted to use IEEE 128-bit floating
   1859  1.10    mrg point, it is not recommended to use
   1860  1.20    mrg <samp class="option">--with-long-double-format=ieee</samp>.
   1861  1.10    mrg </p>
   1862  1.10    mrg </dd>
   1863  1.20    mrg <dt><code class="code">--enable-fdpic</code></dt>
   1864  1.10    mrg <dd><p>On SH Linux systems, generate ELF FDPIC code.
   1865  1.10    mrg </p>
   1866  1.10    mrg </dd>
   1867  1.20    mrg <dt><code class="code">--with-gmp=<var class="var">pathname</var></code></dt>
   1868  1.20    mrg <dt><code class="code">--with-gmp-include=<var class="var">pathname</var></code></dt>
   1869  1.20    mrg <dt><code class="code">--with-gmp-lib=<var class="var">pathname</var></code></dt>
   1870  1.20    mrg <dt><code class="code">--with-mpfr=<var class="var">pathname</var></code></dt>
   1871  1.20    mrg <dt><code class="code">--with-mpfr-include=<var class="var">pathname</var></code></dt>
   1872  1.20    mrg <dt><code class="code">--with-mpfr-lib=<var class="var">pathname</var></code></dt>
   1873  1.20    mrg <dt><code class="code">--with-mpc=<var class="var">pathname</var></code></dt>
   1874  1.20    mrg <dt><code class="code">--with-mpc-include=<var class="var">pathname</var></code></dt>
   1875  1.20    mrg <dt><code class="code">--with-mpc-lib=<var class="var">pathname</var></code></dt>
   1876  1.10    mrg <dd><p>If you want to build GCC but do not have the GMP library, the MPFR
   1877   1.1    mrg library and/or the MPC library installed in a standard location and
   1878   1.3  skrll do not have their sources present in the GCC source tree then you
   1879   1.3  skrll can explicitly specify the directory where they are installed
   1880  1.20    mrg (&lsquo;<samp class="samp">--with-gmp=<var class="var">gmpinstalldir</var></samp>&rsquo;,
   1881  1.20    mrg &lsquo;<samp class="samp">--with-mpfr=<var class="var">mpfrinstalldir</var></samp>&rsquo;,
   1882  1.20    mrg &lsquo;<samp class="samp">--with-mpc=<var class="var">mpcinstalldir</var></samp>&rsquo;).  The
   1883  1.20    mrg <samp class="option">--with-gmp=<var class="var">gmpinstalldir</var></samp> option is shorthand for
   1884  1.20    mrg <samp class="option">--with-gmp-lib=<var class="var">gmpinstalldir</var>/lib</samp> and
   1885  1.20    mrg <samp class="option">--with-gmp-include=<var class="var">gmpinstalldir</var>/include</samp>.  Likewise the
   1886  1.20    mrg <samp class="option">--with-mpfr=<var class="var">mpfrinstalldir</var></samp> option is shorthand for
   1887  1.20    mrg <samp class="option">--with-mpfr-lib=<var class="var">mpfrinstalldir</var>/lib</samp> and
   1888  1.20    mrg <samp class="option">--with-mpfr-include=<var class="var">mpfrinstalldir</var>/include</samp>, also the
   1889  1.20    mrg <samp class="option">--with-mpc=<var class="var">mpcinstalldir</var></samp> option is shorthand for
   1890  1.20    mrg <samp class="option">--with-mpc-lib=<var class="var">mpcinstalldir</var>/lib</samp> and
   1891  1.20    mrg <samp class="option">--with-mpc-include=<var class="var">mpcinstalldir</var>/include</samp>.  If these
   1892   1.1    mrg shorthand assumptions are not correct, you can use the explicit
   1893   1.3  skrll include and lib options directly.  You might also need to ensure the
   1894   1.3  skrll shared libraries can be found by the dynamic linker when building and
   1895   1.3  skrll using GCC, for example by setting the runtime shared library path
   1896  1.20    mrg variable (<code class="env">LD_LIBRARY_PATH</code> on GNU/Linux and Solaris systems).
   1897  1.10    mrg </p>
   1898  1.10    mrg <p>These flags are applicable to the host platform only.  When building
   1899   1.3  skrll a cross compiler, they will not be used to configure target libraries.
   1900  1.10    mrg </p>
   1901  1.10    mrg </dd>
   1902  1.20    mrg <dt><code class="code">--with-isl=<var class="var">pathname</var></code></dt>
   1903  1.20    mrg <dt><code class="code">--with-isl-include=<var class="var">pathname</var></code></dt>
   1904  1.20    mrg <dt><code class="code">--with-isl-lib=<var class="var">pathname</var></code></dt>
   1905  1.10    mrg <dd><p>If you do not have the isl library installed in a standard location and you
   1906   1.5    mrg want to build GCC, you can explicitly specify the directory where it is
   1907  1.20    mrg installed (&lsquo;<samp class="samp">--with-isl=<var class="var">islinstalldir</var></samp>&rsquo;). The
   1908  1.20    mrg <samp class="option">--with-isl=<var class="var">islinstalldir</var></samp> option is shorthand for
   1909  1.20    mrg <samp class="option">--with-isl-lib=<var class="var">islinstalldir</var>/lib</samp> and
   1910  1.20    mrg <samp class="option">--with-isl-include=<var class="var">islinstalldir</var>/include</samp>. If this
   1911   1.5    mrg shorthand assumption is not correct, you can use the explicit
   1912   1.1    mrg include and lib options directly.
   1913  1.10    mrg </p>
   1914  1.10    mrg <p>These flags are applicable to the host platform only.  When building
   1915   1.3  skrll a cross compiler, they will not be used to configure target libraries.
   1916  1.10    mrg </p>
   1917  1.10    mrg </dd>
   1918  1.20    mrg <dt><code class="code">--with-stage1-ldflags=<var class="var">flags</var></code></dt>
   1919  1.10    mrg <dd><p>This option may be used to set linker flags to be used when linking
   1920   1.1    mrg stage 1 of GCC.  These are also used when linking GCC if configured with
   1921  1.20    mrg <samp class="option">--disable-bootstrap</samp>.  If <samp class="option">--with-stage1-libs</samp> is not set to a
   1922  1.20    mrg value, then the default is &lsquo;<samp class="samp">-static-libstdc++ -static-libgcc</samp>&rsquo;, if
   1923   1.7    mrg supported.
   1924  1.10    mrg </p>
   1925  1.10    mrg </dd>
   1926  1.20    mrg <dt><code class="code">--with-stage1-libs=<var class="var">libs</var></code></dt>
   1927  1.10    mrg <dd><p>This option may be used to set libraries to be used when linking stage 1
   1928   1.1    mrg of GCC.  These are also used when linking GCC if configured with
   1929  1.20    mrg <samp class="option">--disable-bootstrap</samp>.
   1930  1.10    mrg </p>
   1931  1.10    mrg </dd>
   1932  1.20    mrg <dt><code class="code">--with-boot-ldflags=<var class="var">flags</var></code></dt>
   1933  1.10    mrg <dd><p>This option may be used to set linker flags to be used when linking
   1934   1.7    mrg stage 2 and later when bootstrapping GCC.  If &ndash;with-boot-libs
   1935   1.7    mrg is not is set to a value, then the default is
   1936  1.20    mrg &lsquo;<samp class="samp">-static-libstdc++ -static-libgcc</samp>&rsquo;.
   1937  1.10    mrg </p>
   1938  1.10    mrg </dd>
   1939  1.20    mrg <dt><code class="code">--with-boot-libs=<var class="var">libs</var></code></dt>
   1940  1.10    mrg <dd><p>This option may be used to set libraries to be used when linking stage 2
   1941   1.7    mrg and later when bootstrapping GCC.
   1942  1.10    mrg </p>
   1943  1.10    mrg </dd>
   1944  1.20    mrg <dt><code class="code">--with-debug-prefix-map=<var class="var">map</var></code></dt>
   1945  1.20    mrg <dd><p>Convert source directory names using <samp class="option">-fdebug-prefix-map</samp> when
   1946  1.20    mrg building runtime libraries.  &lsquo;<samp class="samp"><var class="var">map</var></samp>&rsquo; is a space-separated
   1947  1.20    mrg list of maps of the form &lsquo;<samp class="samp"><var class="var">old</var>=<var class="var">new</var></samp>&rsquo;.
   1948  1.10    mrg </p>
   1949  1.10    mrg </dd>
   1950  1.20    mrg <dt><code class="code">--enable-linker-build-id</code></dt>
   1951  1.20    mrg <dd><p>Tells GCC to pass <samp class="option">--build-id</samp> option to the linker for all final
   1952  1.20    mrg links (links performed without the <samp class="option">-r</samp> or <samp class="option">--relocatable</samp>
   1953   1.1    mrg option), if the linker supports it.  If you specify
   1954  1.20    mrg <samp class="option">--enable-linker-build-id</samp>, but your linker does not
   1955  1.20    mrg support <samp class="option">--build-id</samp> option, a warning is issued and the
   1956  1.20    mrg <samp class="option">--enable-linker-build-id</samp> option is ignored.  The default is off.
   1957  1.10    mrg </p>
   1958  1.10    mrg </dd>
   1959  1.20    mrg <dt><code class="code">--with-linker-hash-style=<var class="var">choice</var></code></dt>
   1960  1.20    mrg <dd><p>Tells GCC to pass <samp class="option">--hash-style=<var class="var">choice</var></samp> option to the
   1961  1.20    mrg linker for all final links. <var class="var">choice</var> can be one of
   1962  1.20    mrg &lsquo;<samp class="samp">sysv</samp>&rsquo;, &lsquo;<samp class="samp">gnu</samp>&rsquo;, and &lsquo;<samp class="samp">both</samp>&rsquo; where &lsquo;<samp class="samp">sysv</samp>&rsquo; is the default.
   1963  1.10    mrg </p>
   1964  1.10    mrg </dd>
   1965  1.20    mrg <dt><code class="code">--enable-gnu-unique-object</code></dt>
   1966  1.20    mrg <dt><code class="code">--disable-gnu-unique-object</code></dt>
   1967  1.10    mrg <dd><p>Tells GCC to use the gnu_unique_object relocation for C++ template
   1968   1.1    mrg static data members and inline function local statics.  Enabled by
   1969   1.5    mrg default for a toolchain with an assembler that accepts it and
   1970   1.1    mrg GLIBC 2.11 or above, otherwise disabled.
   1971  1.10    mrg </p>
   1972  1.10    mrg </dd>
   1973  1.20    mrg <dt><code class="code">--with-diagnostics-color=<var class="var">choice</var></code></dt>
   1974  1.20    mrg <dd><p>Tells GCC to use <var class="var">choice</var> as the default for <samp class="option">-fdiagnostics-color=</samp>
   1975  1.20    mrg option (if not used explicitly on the command line).  <var class="var">choice</var>
   1976  1.20    mrg can be one of &lsquo;<samp class="samp">never</samp>&rsquo;, &lsquo;<samp class="samp">auto</samp>&rsquo;, &lsquo;<samp class="samp">always</samp>&rsquo;, and &lsquo;<samp class="samp">auto-if-env</samp>&rsquo;
   1977  1.20    mrg where &lsquo;<samp class="samp">auto</samp>&rsquo; is the default.  &lsquo;<samp class="samp">auto-if-env</samp>&rsquo; makes
   1978  1.20    mrg <samp class="option">-fdiagnostics-color=auto</samp> the default if <code class="env">GCC_COLORS</code>
   1979  1.13    mrg is present and non-empty in the environment of the compiler, and
   1980  1.20    mrg <samp class="option">-fdiagnostics-color=never</samp> otherwise.
   1981  1.10    mrg </p>
   1982  1.10    mrg </dd>
   1983  1.20    mrg <dt><code class="code">--with-diagnostics-urls=<var class="var">choice</var></code></dt>
   1984  1.20    mrg <dd><p>Tells GCC to use <var class="var">choice</var> as the default for <samp class="option">-fdiagnostics-urls=</samp>
   1985  1.20    mrg option (if not used explicitly on the command line).  <var class="var">choice</var>
   1986  1.20    mrg can be one of &lsquo;<samp class="samp">never</samp>&rsquo;, &lsquo;<samp class="samp">auto</samp>&rsquo;, &lsquo;<samp class="samp">always</samp>&rsquo;, and &lsquo;<samp class="samp">auto-if-env</samp>&rsquo;
   1987  1.20    mrg where &lsquo;<samp class="samp">auto</samp>&rsquo; is the default.  &lsquo;<samp class="samp">auto-if-env</samp>&rsquo; makes
   1988  1.20    mrg <samp class="option">-fdiagnostics-urls=auto</samp> the default if <code class="env">GCC_URLS</code>
   1989  1.20    mrg or <code class="env">TERM_URLS</code> is present and non-empty in the environment of the
   1990  1.20    mrg compiler, and <samp class="option">-fdiagnostics-urls=never</samp> otherwise.
   1991  1.13    mrg </p>
   1992  1.13    mrg </dd>
   1993  1.20    mrg <dt><code class="code">--enable-lto</code></dt>
   1994  1.20    mrg <dt><code class="code">--disable-lto</code></dt>
   1995  1.10    mrg <dd><p>Enable support for link-time optimization (LTO).  This is enabled by
   1996  1.20    mrg default, and may be disabled using <samp class="option">--disable-lto</samp>.
   1997  1.10    mrg </p>
   1998  1.10    mrg </dd>
   1999  1.20    mrg <dt><code class="code">--enable-linker-plugin-configure-flags=FLAGS</code></dt>
   2000  1.20    mrg <dt><code class="code">--enable-linker-plugin-flags=FLAGS</code></dt>
   2001  1.10    mrg <dd><p>By default, linker plugins (such as the LTO plugin) are built for the
   2002   1.5    mrg host system architecture.  For the case that the linker has a
   2003   1.5    mrg different (but run-time compatible) architecture, these flags can be
   2004   1.5    mrg specified to build plugins that are compatible to the linker.  For
   2005   1.5    mrg example, if you are building GCC for a 64-bit x86_64
   2006  1.20    mrg (&lsquo;<samp class="samp">x86_64-pc-linux-gnu</samp>&rsquo;) host system, but have a 32-bit x86
   2007  1.20    mrg GNU/Linux (&lsquo;<samp class="samp">i686-pc-linux-gnu</samp>&rsquo;) linker executable (which is
   2008   1.5    mrg executable on the former system), you can configure GCC as follows for
   2009   1.5    mrg getting compatible linker plugins:
   2010  1.10    mrg </p>
   2011  1.20    mrg <div class="example smallexample">
   2012  1.20    mrg <pre class="example-preformatted">% <var class="var">srcdir</var>/configure \
   2013  1.12    mrg     --host=x86_64-pc-linux-gnu \
   2014  1.10    mrg     --enable-linker-plugin-configure-flags=--host=i686-pc-linux-gnu \
   2015  1.10    mrg     --enable-linker-plugin-flags='CC=gcc\ -m32\ -Wl,-rpath,[...]/i686-pc-linux-gnu/lib'
   2016  1.10    mrg </pre></div>
   2017  1.10    mrg 
   2018  1.10    mrg </dd>
   2019  1.20    mrg <dt><code class="code">--with-plugin-ld=<var class="var">pathname</var></code></dt>
   2020  1.10    mrg <dd><p>Enable an alternate linker to be used at link-time optimization (LTO)
   2021  1.20    mrg link time when <samp class="option">-fuse-linker-plugin</samp> is enabled.
   2022   1.3  skrll This linker should have plugin support such as gold starting with
   2023  1.10    mrg version 2.20 or GNU ld starting with version 2.21.
   2024  1.20    mrg See <samp class="option">-fuse-linker-plugin</samp> for details.
   2025  1.10    mrg </p>
   2026  1.10    mrg </dd>
   2027  1.20    mrg <dt><code class="code">--enable-canonical-system-headers</code></dt>
   2028  1.20    mrg <dt><code class="code">--disable-canonical-system-headers</code></dt>
   2029  1.20    mrg <dd><p>Enable system header path canonicalization for <samp class="file">libcpp</samp>.  This can
   2030   1.3  skrll produce shorter header file paths in diagnostics and dependency output
   2031   1.3  skrll files, but these changed header paths may conflict with some compilation
   2032   1.3  skrll environments.  Enabled by default, and may be disabled using
   2033  1.20    mrg <samp class="option">--disable-canonical-system-headers</samp>.
   2034  1.10    mrg </p>
   2035  1.10    mrg </dd>
   2036  1.20    mrg <dt><code class="code">--with-glibc-version=<var class="var">major</var>.<var class="var">minor</var></code></dt>
   2037  1.10    mrg <dd><p>Tell GCC that when the GNU C Library (glibc) is used on the target it
   2038  1.20    mrg will be version <var class="var">major</var>.<var class="var">minor</var> or later.  Normally this can
   2039  1.10    mrg be detected from the C library&rsquo;s header files, but this option may be
   2040   1.5    mrg needed when bootstrapping a cross toolchain without the header files
   2041   1.5    mrg available for building the initial bootstrap compiler.
   2042  1.10    mrg </p>
   2043  1.10    mrg <p>If GCC is configured with some multilibs that use glibc and some that
   2044  1.10    mrg do not, this option applies only to the multilibs that use glibc.
   2045   1.5    mrg However, such configurations may not work well as not all the relevant
   2046   1.5    mrg configuration in GCC is on a per-multilib basis.
   2047  1.10    mrg </p>
   2048  1.10    mrg </dd>
   2049  1.20    mrg <dt><code class="code">--enable-as-accelerator-for=<var class="var">target</var></code></dt>
   2050  1.20    mrg <dd><p>Build as offload target compiler. Specify offload host triple by <var class="var">target</var>.
   2051  1.10    mrg </p>
   2052  1.10    mrg </dd>
   2053  1.20    mrg <dt><code class="code">--enable-offload-targets=<var class="var">target1</var>[=<var class="var">path1</var>],&hellip;,<var class="var">targetN</var>[=<var class="var">pathN</var>]</code></dt>
   2054  1.20    mrg <dd><p>Enable offloading to targets <var class="var">target1</var>, &hellip;, <var class="var">targetN</var>.
   2055   1.5    mrg Offload compilers are expected to be already installed.  Default search
   2056  1.20    mrg path for them is <samp class="file"><var class="var">exec-prefix</var></samp>, but it can be changed by
   2057  1.20    mrg specifying paths <var class="var">path1</var>, &hellip;, <var class="var">pathN</var>.
   2058  1.10    mrg </p>
   2059  1.20    mrg <div class="example smallexample">
   2060  1.20    mrg <pre class="example-preformatted">% <var class="var">srcdir</var>/configure \
   2061  1.19    mrg     --enable-offload-targets=amdgcn-amdhsa,nvptx-none
   2062  1.10    mrg </pre></div>
   2063   1.7    mrg 
   2064  1.16    mrg </dd>
   2065  1.20    mrg <dt><code class="code">--enable-offload-defaulted</code></dt>
   2066  1.16    mrg <dd>
   2067  1.16    mrg <p>Tell GCC that configured but not installed offload compilers and libgomp
   2068  1.16    mrg plugins are silently ignored.  Useful for distribution compilers where
   2069  1.16    mrg those are in separate optional packages and where the presence or absence
   2070  1.16    mrg of those optional packages should determine the actual supported offloading
   2071  1.16    mrg target set rather than the GCC configure-time selection.
   2072  1.10    mrg </p>
   2073  1.10    mrg </dd>
   2074  1.20    mrg <dt><code class="code">--enable-cet</code></dt>
   2075  1.20    mrg <dt><code class="code">--disable-cet</code></dt>
   2076  1.10    mrg <dd><p>Enable building target run-time libraries with control-flow
   2077  1.20    mrg instrumentation, see <samp class="option">-fcf-protection</samp> option.  When
   2078  1.20    mrg <code class="code">--enable-cet</code> is specified target libraries are configured
   2079  1.20    mrg to add <samp class="option">-fcf-protection</samp> and, if needed, other target
   2080  1.10    mrg specific options to a set of building options.
   2081  1.10    mrg </p>
   2082  1.20    mrg <p><code class="code">--enable-cet=auto</code> is default.  CET is enabled on Linux/x86 if
   2083  1.20    mrg target binutils supports <code class="code">Intel CET</code> instructions and disabled
   2084  1.16    mrg otherwise.  In this case, the target libraries are configured to get
   2085  1.20    mrg additional <samp class="option">-fcf-protection</samp> option.
   2086  1.12    mrg </p>
   2087  1.12    mrg </dd>
   2088  1.20    mrg <dt><code class="code">--with-riscv-attribute=&lsquo;<samp class="samp">yes</samp>&rsquo;, &lsquo;<samp class="samp">no</samp>&rsquo; or &lsquo;<samp class="samp">default</samp>&rsquo;</code></dt>
   2089  1.12    mrg <dd><p>Generate RISC-V attribute by default, in order to record extra build
   2090  1.12    mrg information in object.
   2091  1.12    mrg </p>
   2092  1.12    mrg <p>The option is disabled by default. It is enabled on RISC-V/ELF (bare-metal)
   2093  1.12    mrg target if target binutils supported.
   2094  1.16    mrg </p>
   2095  1.16    mrg </dd>
   2096  1.20    mrg <dt><code class="code">--enable-s390-excess-float-precision</code></dt>
   2097  1.20    mrg <dt><code class="code">--disable-s390-excess-float-precision</code></dt>
   2098  1.16    mrg <dd><p>On s390(x) targets, enable treatment of float expressions with double precision
   2099  1.20    mrg when in standards-compliant mode (e.g., when <code class="code">--std=c99</code> or
   2100  1.20    mrg <code class="code">-fexcess-precision=standard</code> are given).
   2101  1.16    mrg </p>
   2102  1.16    mrg <p>For a native build and cross compiles that have target headers, the option&rsquo;s
   2103  1.16    mrg default is derived from glibc&rsquo;s behavior. When glibc clamps float_t to double,
   2104  1.16    mrg GCC follows and enables the option. For other cross compiles, the default is
   2105  1.16    mrg disabled.
   2106  1.16    mrg </p>
   2107  1.16    mrg </dd>
   2108  1.20    mrg <dt><code class="code">--with-zstd=<var class="var">pathname</var></code></dt>
   2109  1.20    mrg <dt><code class="code">--with-zstd-include=<var class="var">pathname</var></code></dt>
   2110  1.20    mrg <dt><code class="code">--with-zstd-lib=<var class="var">pathname</var></code></dt>
   2111  1.20    mrg <dd><p>If you do not have the <code class="code">zstd</code> library installed in a standard
   2112  1.16    mrg location and you want to build GCC, you can explicitly specify the
   2113  1.20    mrg directory where it is installed (&lsquo;<samp class="samp">--with-zstd=<var class="var">zstdinstalldir</var></samp>&rsquo;).
   2114  1.20    mrg The <samp class="option">--with-zstd=<var class="var">zstdinstalldir</var></samp> option is shorthand for
   2115  1.20    mrg <samp class="option">--with-zstd-lib=<var class="var">zstdinstalldir</var>/lib</samp> and
   2116  1.20    mrg <samp class="option">--with-zstd-include=<var class="var">zstdinstalldir</var>/include</samp>. If this
   2117  1.16    mrg shorthand assumption is not correct, you can use the explicit
   2118  1.16    mrg include and lib options directly.
   2119  1.16    mrg </p>
   2120  1.16    mrg <p>These flags are applicable to the host platform only.  When building
   2121  1.16    mrg a cross compiler, they will not be used to configure target libraries.
   2122  1.10    mrg </p></dd>
   2123   1.6    mrg </dl>
   2124   1.1    mrg 
   2125  1.20    mrg <h4 class="subheading" id="Cross-Compiler-Specific-Options"><span>Cross-Compiler-Specific Options<a class="copiable-link" href="#Cross-Compiler-Specific-Options"> &para;</a></span></h4>
   2126   1.1    mrg <p>The following options only apply to building cross compilers.
   2127  1.10    mrg </p>
   2128  1.20    mrg <dl class="table">
   2129  1.20    mrg <dt><code class="code">--with-toolexeclibdir=<var class="var">dir</var></code></dt>
   2130  1.13    mrg <dd><p>Specify the installation directory for libraries built with a cross compiler.
   2131  1.20    mrg The default is <samp class="option">${gcc_tooldir}/lib</samp>.
   2132  1.13    mrg </p>
   2133  1.13    mrg </dd>
   2134  1.20    mrg <dt><code class="code">--with-sysroot</code></dt>
   2135  1.20    mrg <dt><code class="code">--with-sysroot=<var class="var">dir</var></code></dt>
   2136  1.20    mrg <dd><p>Tells GCC to consider <var class="var">dir</var> as the root of a tree that contains
   2137  1.10    mrg (a subset of) the root filesystem of the target operating system.
   2138   1.1    mrg Target system headers, libraries and run-time object files will be
   2139   1.3  skrll searched for in there.  More specifically, this acts as if
   2140  1.20    mrg <samp class="option">--sysroot=<var class="var">dir</var></samp> was added to the default options of the built
   2141   1.1    mrg compiler.  The specified directory is not copied into the
   2142  1.20    mrg install tree, unlike the options <samp class="option">--with-headers</samp> and
   2143  1.20    mrg <samp class="option">--with-libs</samp> that this option obsoletes.  The default value,
   2144  1.20    mrg in case <samp class="option">--with-sysroot</samp> is not given an argument, is
   2145  1.20    mrg <samp class="option">${gcc_tooldir}/sys-root</samp>.  If the specified directory is a
   2146  1.20    mrg subdirectory of <samp class="option">${exec_prefix}</samp>, then it will be found relative to
   2147   1.1    mrg the GCC binaries if the installation tree is moved.
   2148  1.10    mrg </p>
   2149  1.10    mrg <p>This option affects the system root for the compiler used to build
   2150   1.1    mrg target libraries (which runs on the build system) and the compiler newly
   2151  1.20    mrg installed with <code class="code">make install</code>; it does not affect the compiler which is
   2152   1.1    mrg used to build GCC itself.
   2153  1.10    mrg </p>
   2154  1.20    mrg <p>If you specify the <samp class="option">--with-native-system-header-dir=<var class="var">dirname</var></samp>
   2155  1.20    mrg option then the compiler will search that directory within <var class="var">dirname</var> for
   2156  1.20    mrg native system headers rather than the default <samp class="file">/usr/include</samp>.
   2157  1.10    mrg </p>
   2158  1.10    mrg </dd>
   2159  1.20    mrg <dt><code class="code">--with-build-sysroot</code></dt>
   2160  1.20    mrg <dt><code class="code">--with-build-sysroot=<var class="var">dir</var></code></dt>
   2161  1.20    mrg <dd><p>Tells GCC to consider <var class="var">dir</var> as the system root (see
   2162  1.20    mrg <samp class="option">--with-sysroot</samp>) while building target libraries, instead of
   2163  1.20    mrg the directory specified with <samp class="option">--with-sysroot</samp>.  This option is
   2164  1.20    mrg only useful when you are already using <samp class="option">--with-sysroot</samp>.  You
   2165  1.20    mrg can use <samp class="option">--with-build-sysroot</samp> when you are configuring with
   2166  1.20    mrg <samp class="option">--prefix</samp> set to a directory that is different from the one in
   2167   1.1    mrg which you are installing GCC and your target libraries.
   2168  1.10    mrg </p>
   2169  1.10    mrg <p>This option affects the system root for the compiler used to build
   2170   1.1    mrg target libraries (which runs on the build system); it does not affect
   2171   1.1    mrg the compiler which is used to build GCC itself.
   2172  1.10    mrg </p>
   2173  1.20    mrg <p>If you specify the <samp class="option">--with-native-system-header-dir=<var class="var">dirname</var></samp>
   2174  1.20    mrg option then the compiler will search that directory within <var class="var">dirname</var> for
   2175  1.20    mrg native system headers rather than the default <samp class="file">/usr/include</samp>.
   2176  1.10    mrg </p>
   2177  1.10    mrg </dd>
   2178  1.20    mrg <dt><code class="code">--with-headers</code></dt>
   2179  1.20    mrg <dt><code class="code">--with-headers=<var class="var">dir</var></code></dt>
   2180  1.20    mrg <dd><p>Deprecated in favor of <samp class="option">--with-sysroot</samp>.
   2181  1.10    mrg Specifies that target headers are available when building a cross compiler.
   2182  1.20    mrg The <var class="var">dir</var> argument specifies a directory which has the target include
   2183  1.20    mrg files.  These include files will be copied into the <samp class="file">gcc</samp> install
   2184  1.20    mrg directory.  <em class="emph">This option with the <var class="var">dir</var> argument is required</em> when
   2185  1.20    mrg building a cross compiler, if <samp class="file"><var class="var">prefix</var>/<var class="var">target</var>/sys-include</samp>
   2186  1.20    mrg doesn&rsquo;t pre-exist.  If <samp class="file"><var class="var">prefix</var>/<var class="var">target</var>/sys-include</samp> does
   2187  1.20    mrg pre-exist, the <var class="var">dir</var> argument may be omitted.  <code class="command">fixincludes</code>
   2188   1.1    mrg will be run on these files to make them compatible with GCC.
   2189  1.10    mrg </p>
   2190  1.10    mrg </dd>
   2191  1.20    mrg <dt><code class="code">--without-headers</code></dt>
   2192  1.10    mrg <dd><p>Tells GCC not use any target headers from a libc when building a cross
   2193   1.1    mrg compiler.  When crossing to GNU/Linux, you need the headers so GCC
   2194   1.1    mrg can build the exception handling for libgcc.
   2195  1.10    mrg </p>
   2196  1.10    mrg </dd>
   2197  1.20    mrg <dt><code class="code">--with-libs</code></dt>
   2198  1.20    mrg <dt><code class="code">--with-libs=&quot;<var class="var">dir1</var> <var class="var">dir2</var> &hellip; <var class="var">dirN</var>&quot;</code></dt>
   2199  1.20    mrg <dd><p>Deprecated in favor of <samp class="option">--with-sysroot</samp>.
   2200   1.1    mrg Specifies a list of directories which contain the target runtime
   2201  1.20    mrg libraries.  These libraries will be copied into the <samp class="file">gcc</samp> install
   2202   1.1    mrg directory.  If the directory list is omitted, this option has no
   2203   1.1    mrg effect.
   2204  1.10    mrg </p>
   2205  1.10    mrg </dd>
   2206  1.20    mrg <dt><code class="code">--with-newlib</code></dt>
   2207  1.20    mrg <dd><p>Specifies that &lsquo;<samp class="samp">newlib</samp>&rsquo; is
   2208  1.20    mrg being used as the target C library.  This causes <code class="code">__eprintf</code> to be
   2209  1.20    mrg omitted from <samp class="file">libgcc.a</samp> on the assumption that it will be provided by
   2210  1.20    mrg &lsquo;<samp class="samp">newlib</samp>&rsquo;.
   2211  1.10    mrg </p>
   2212  1.13    mrg <a name="avr"></a>
   2213  1.10    mrg </dd>
   2214  1.20    mrg <dt><code class="code">--with-avrlibc</code></dt>
   2215  1.20    mrg <dd><p>Only supported for the AVR target. Specifies that &lsquo;<samp class="samp">AVR-Libc</samp>&rsquo; is
   2216  1.13    mrg being used as the target C&nbsp; library.  This causes float support
   2217  1.20    mrg functions like <code class="code">__addsf3</code> to be omitted from <samp class="file">libgcc.a</samp> on
   2218  1.20    mrg the assumption that it will be provided by <samp class="file">libm.a</samp>.  For more
   2219  1.20    mrg technical details, cf. <a class="uref" href="https://gcc.gnu.org/PR54461">PR54461</a>.
   2220  1.13    mrg It is not supported for
   2221   1.3  skrll RTEMS configurations, which currently use newlib.  The option is
   2222   1.3  skrll supported since version 4.7.2 and is the default in 4.8.0 and newer.
   2223  1.10    mrg </p>
   2224  1.10    mrg </dd>
   2225  1.20    mrg <dt><code class="code">--with-double={32|64|32,64|64,32}</code></dt>
   2226  1.20    mrg <dt><code class="code">--with-long-double={32|64|32,64|64,32|double}</code></dt>
   2227  1.13    mrg <dd><p>Only supported for the AVR target since version&nbsp;10.
   2228  1.20    mrg Specify the default layout available for the C/C++ &lsquo;<samp class="samp">double</samp>&rsquo;
   2229  1.20    mrg and &lsquo;<samp class="samp">long double</samp>&rsquo; type, respectively. The following rules apply:
   2230  1.20    mrg </p><ul class="itemize mark-bullet">
   2231  1.20    mrg <li>The first value after the &lsquo;<samp class="samp">=</samp>&rsquo; specifies the default layout (in bits)
   2232  1.20    mrg of the type and also the default for the <samp class="option">-mdouble=</samp> resp.
   2233  1.20    mrg <samp class="option">-mlong-double=</samp> compiler option.
   2234  1.20    mrg </li><li>If more than one value is specified, respective multilib variants are
   2235  1.20    mrg available, and  <samp class="option">-mdouble=</samp> resp. <samp class="option">-mlong-double=</samp> acts
   2236  1.13    mrg as a multilib option.
   2237  1.20    mrg </li><li>If <samp class="option">--with-long-double=double</samp> is specified, &lsquo;<samp class="samp">double</samp>&rsquo; and
   2238  1.20    mrg &lsquo;<samp class="samp">long double</samp>&rsquo; will have the same layout.
   2239  1.20    mrg </li><li>The defaults are <samp class="option">--with-long-double=64,32</samp> and
   2240  1.20    mrg <samp class="option">--with-double=32,64</samp>.  The default &lsquo;<samp class="samp">double</samp>&rsquo; layout imposed by
   2241  1.13    mrg the latter is compatible with older versions of the compiler that implement
   2242  1.20    mrg &lsquo;<samp class="samp">double</samp>&rsquo; as a 32-bit type, which does not comply to the language standard.
   2243  1.13    mrg </li></ul>
   2244  1.20    mrg <p>Not all combinations of <samp class="option">--with-double=</samp> and
   2245  1.20    mrg <samp class="option">--with-long-double=</samp> are valid.  For example, the combination
   2246  1.20    mrg <samp class="option">--with-double=32,64</samp> <samp class="option">--with-long-double=32</samp> will be
   2247  1.13    mrg rejected because the first option specifies the availability of
   2248  1.20    mrg multilibs for &lsquo;<samp class="samp">double</samp>&rsquo;, whereas the second option implies
   2249  1.20    mrg that &lsquo;<samp class="samp">long double</samp>&rsquo; &mdash; and hence also &lsquo;<samp class="samp">double</samp>&rsquo; &mdash; is always
   2250  1.13    mrg 32&nbsp;bits wide.
   2251  1.13    mrg </p>
   2252  1.13    mrg </dd>
   2253  1.20    mrg <dt><code class="code">--with-double-comparison={tristate|bool|libf7}</code></dt>
   2254  1.13    mrg <dd><p>Only supported for the AVR target since version&nbsp;10.
   2255  1.13    mrg Specify what result format is returned by library functions that
   2256  1.20    mrg compare 64-bit floating point values (<code class="code">DFmode</code>).
   2257  1.20    mrg The GCC default is &lsquo;<samp class="samp">tristate</samp>&rsquo;.  If the floating point
   2258  1.20    mrg implementation returns a boolean instead, set it to &lsquo;<samp class="samp">bool</samp>&rsquo;.
   2259  1.13    mrg </p>
   2260  1.13    mrg </dd>
   2261  1.20    mrg <dt><code class="code">--with-libf7={libgcc|math|math-symbols|no}</code></dt>
   2262  1.13    mrg <dd><p>Only supported for the AVR target since version&nbsp;10.
   2263  1.13    mrg Specify to which degree code from LibF7 is included in libgcc.
   2264  1.13    mrg LibF7 is an ad-hoc, AVR-specific, 64-bit floating point emulation
   2265  1.20    mrg written in C and (inline) assembly. &lsquo;<samp class="samp">libgcc</samp>&rsquo; adds support
   2266  1.13    mrg for functions that one would usually expect in libgcc like double addition,
   2267  1.20    mrg double comparisons and double conversions. &lsquo;<samp class="samp">math</samp>&rsquo; also adds routines
   2268  1.20    mrg that one would expect in <samp class="file">libm.a</samp>, but with <code class="code">__</code> (two underscores)
   2269  1.20    mrg prepended to the symbol names as specified by <samp class="file">math.h</samp>.
   2270  1.20    mrg &lsquo;<samp class="samp">math-symbols</samp>&rsquo; also defines weak aliases for the functions
   2271  1.20    mrg declared in <samp class="file">math.h</samp>.  However, <code class="code">--with-libf7</code> won&rsquo;t
   2272  1.20    mrg install no <samp class="file">math.h</samp> header file whatsoever, this file must come
   2273  1.20    mrg from elsewhere.  This option sets <samp class="option">--with-double-comparison</samp>
   2274  1.20    mrg to &lsquo;<samp class="samp">bool</samp>&rsquo;.
   2275  1.13    mrg </p>
   2276  1.13    mrg </dd>
   2277  1.20    mrg <dt><code class="code">--with-nds32-lib=<var class="var">library</var></code></dt>
   2278  1.20    mrg <dd><p>Specifies that <var class="var">library</var> setting is used for building <samp class="file">libgcc.a</samp>.
   2279  1.20    mrg Currently, the valid <var class="var">library</var> is &lsquo;<samp class="samp">newlib</samp>&rsquo; or &lsquo;<samp class="samp">mculib</samp>&rsquo;.
   2280   1.5    mrg This option is only supported for the NDS32 target.
   2281  1.10    mrg </p>
   2282  1.10    mrg </dd>
   2283  1.20    mrg <dt><code class="code">--with-build-time-tools=<var class="var">dir</var></code></dt>
   2284  1.10    mrg <dd><p>Specifies where to find the set of target tools (assembler, linker, etc.)
   2285   1.1    mrg that will be used while building GCC itself.  This option can be useful
   2286   1.1    mrg if the directory layouts are different between the system you are building
   2287   1.1    mrg GCC on, and the system where you will deploy it.
   2288  1.10    mrg </p>
   2289  1.20    mrg <p>For example, on an &lsquo;<samp class="samp">ia64-hp-hpux</samp>&rsquo; system, you may have the GNU
   2290  1.20    mrg assembler and linker in <samp class="file">/usr/bin</samp>, and the native tools in a
   2291   1.6    mrg different path, and build a toolchain that expects to find the
   2292  1.20    mrg native tools in <samp class="file">/usr/bin</samp>.
   2293  1.10    mrg </p>
   2294  1.20    mrg <p>When you use this option, you should ensure that <var class="var">dir</var> includes
   2295  1.20    mrg <code class="command">ar</code>, <code class="command">as</code>, <code class="command">ld</code>, <code class="command">nm</code>,
   2296  1.20    mrg <code class="command">ranlib</code> and <code class="command">strip</code> if necessary, and possibly
   2297  1.20    mrg <code class="command">objdump</code>.  Otherwise, GCC may use an inconsistent set of
   2298  1.10    mrg tools.
   2299  1.10    mrg </p></dd>
   2300   1.6    mrg </dl>
   2301   1.1    mrg 
   2302  1.20    mrg <h4 class="subsubheading" id="Overriding-configure-test-results"><span>Overriding <code class="command">configure</code> test results<a class="copiable-link" href="#Overriding-configure-test-results"> &para;</a></span></h4>
   2303   1.1    mrg 
   2304   1.6    mrg <p>Sometimes, it might be necessary to override the result of some
   2305  1.20    mrg <code class="command">configure</code> test, for example in order to ease porting to a new
   2306  1.20    mrg system or work around a bug in a test.  The toplevel <code class="command">configure</code>
   2307   1.6    mrg script provides three variables for this:
   2308  1.10    mrg </p>
   2309  1.20    mrg <dl class="table">
   2310  1.20    mrg <dt><a id="index-build_005fconfigargs"></a><span><code class="code">build_configargs</code><a class="copiable-link" href="#index-build_005fconfigargs"> &para;</a></span></dt>
   2311  1.20    mrg <dd><p>The contents of this variable is passed to all build <code class="command">configure</code>
   2312   1.6    mrg scripts.
   2313  1.10    mrg </p>
   2314  1.10    mrg </dd>
   2315  1.20    mrg <dt><a id="index-host_005fconfigargs"></a><span><code class="code">host_configargs</code><a class="copiable-link" href="#index-host_005fconfigargs"> &para;</a></span></dt>
   2316  1.20    mrg <dd><p>The contents of this variable is passed to all host <code class="command">configure</code>
   2317   1.6    mrg scripts.
   2318  1.10    mrg </p>
   2319  1.10    mrg </dd>
   2320  1.20    mrg <dt><a id="index-target_005fconfigargs"></a><span><code class="code">target_configargs</code><a class="copiable-link" href="#index-target_005fconfigargs"> &para;</a></span></dt>
   2321  1.20    mrg <dd><p>The contents of this variable is passed to all target <code class="command">configure</code>
   2322   1.6    mrg scripts.
   2323  1.10    mrg </p>
   2324  1.10    mrg </dd>
   2325  1.10    mrg </dl>
   2326   1.1    mrg 
   2327  1.20    mrg <p>In order to avoid shell and <code class="command">make</code> quoting issues for complex
   2328  1.20    mrg overrides, you can pass a setting for <code class="env">CONFIG_SITE</code> and set
   2329   1.6    mrg variables in the site file.
   2330  1.10    mrg </p>
   2331  1.20    mrg <h4 class="subheading" id="Objective-C-Specific-Options"><span>Objective-C-Specific Options<a class="copiable-link" href="#Objective-C-Specific-Options"> &para;</a></span></h4>
   2332   1.7    mrg 
   2333   1.9    mrg <p>The following options apply to the build of the Objective-C runtime library.
   2334  1.10    mrg </p>
   2335  1.20    mrg <dl class="table">
   2336  1.20    mrg <dt><code class="code">--enable-objc-gc</code></dt>
   2337  1.10    mrg <dd><p>Specify that an additional variant of the GNU Objective-C runtime library
   2338   1.9    mrg is built, using an external build of the Boehm-Demers-Weiser garbage
   2339  1.20    mrg collector (<a class="uref" href="https://www.hboehm.info/gc/">https://www.hboehm.info/gc/</a>).  This library needs to be
   2340   1.9    mrg available for each multilib variant, unless configured with
   2341  1.20    mrg <samp class="option">--enable-objc-gc=&lsquo;<samp class="samp">auto</samp>&rsquo;</samp> in which case the build of the
   2342   1.9    mrg additional runtime library is skipped when not available and the build
   2343   1.9    mrg continues.
   2344  1.10    mrg </p>
   2345  1.10    mrg </dd>
   2346  1.20    mrg <dt><code class="code">--with-target-bdw-gc=<var class="var">list</var></code></dt>
   2347  1.20    mrg <dt><code class="code">--with-target-bdw-gc-include=<var class="var">list</var></code></dt>
   2348  1.20    mrg <dt><code class="code">--with-target-bdw-gc-lib=<var class="var">list</var></code></dt>
   2349  1.10    mrg <dd><p>Specify search directories for the garbage collector header files and
   2350  1.20    mrg libraries. <var class="var">list</var> is a comma separated list of key value pairs of the
   2351  1.20    mrg form &lsquo;<samp class="samp"><var class="var">multilibdir</var>=<var class="var">path</var></samp>&rsquo;, where the default multilib key
   2352  1.20    mrg is named as &lsquo;<samp class="samp">.</samp>&rsquo; (dot), or is omitted (e.g.
   2353  1.20    mrg &lsquo;<samp class="samp">--with-target-bdw-gc=/opt/bdw-gc,32=/opt-bdw-gc32</samp>&rsquo;).
   2354  1.10    mrg </p>
   2355  1.20    mrg <p>The options <samp class="option">--with-target-bdw-gc-include</samp> and
   2356  1.20    mrg <samp class="option">--with-target-bdw-gc-lib</samp> must always be specified together
   2357   1.9    mrg for each multilib variant and they take precedence over
   2358  1.20    mrg <samp class="option">--with-target-bdw-gc</samp>.  If <samp class="option">--with-target-bdw-gc-include</samp>
   2359   1.9    mrg is missing values for a multilib, then the value for the default
   2360  1.20    mrg multilib is used (e.g. &lsquo;<samp class="samp">--with-target-bdw-gc-include=/opt/bdw-gc/include</samp>&rsquo;
   2361  1.20    mrg &lsquo;<samp class="samp">--with-target-bdw-gc-lib=/opt/bdw-gc/lib64,32=/opt-bdw-gc/lib32</samp>&rsquo;).
   2362   1.9    mrg If none of these options are specified, the library is assumed in
   2363  1.10    mrg default locations.
   2364  1.10    mrg </p></dd>
   2365   1.6    mrg </dl>
   2366   1.1    mrg 
   2367  1.20    mrg <h4 class="subheading" id="D-Specific-Options"><span>D-Specific Options<a class="copiable-link" href="#D-Specific-Options"> &para;</a></span></h4>
   2368  1.12    mrg 
   2369  1.12    mrg <p>The following options apply to the build of the D runtime library.
   2370  1.12    mrg </p>
   2371  1.20    mrg <dl class="table">
   2372  1.20    mrg <dt><code class="code">--enable-libphobos-checking</code></dt>
   2373  1.20    mrg <dt><code class="code">--disable-libphobos-checking</code></dt>
   2374  1.20    mrg <dt><code class="code">--enable-libphobos-checking=<var class="var">list</var></code></dt>
   2375  1.13    mrg <dd><p>This option controls whether run-time checks and contracts are compiled into
   2376  1.13    mrg the D runtime library.  When the option is not specified, the library is built
   2377  1.20    mrg with &lsquo;<samp class="samp">release</samp>&rsquo; checking.  When the option is specified without a
   2378  1.20    mrg <var class="var">list</var>, the result is the same as &lsquo;<samp class="samp">--enable-libphobos-checking=yes</samp>&rsquo;.
   2379  1.20    mrg Likewise, &lsquo;<samp class="samp">--disable-libphobos-checking</samp>&rsquo; is equivalent to
   2380  1.20    mrg &lsquo;<samp class="samp">--enable-libphobos-checking=no</samp>&rsquo;.
   2381  1.20    mrg </p>
   2382  1.20    mrg <p>The categories of checks available in <var class="var">list</var> are &lsquo;<samp class="samp">yes</samp>&rsquo; (compiles
   2383  1.20    mrg libphobos with <samp class="option">-fno-release</samp>), &lsquo;<samp class="samp">no</samp>&rsquo; (compiles libphobos with
   2384  1.20    mrg <samp class="option">-frelease</samp>), &lsquo;<samp class="samp">all</samp>&rsquo; (same as &lsquo;<samp class="samp">yes</samp>&rsquo;), &lsquo;<samp class="samp">none</samp>&rsquo; or
   2385  1.20    mrg &lsquo;<samp class="samp">release</samp>&rsquo; (same as &lsquo;<samp class="samp">no</samp>&rsquo;).
   2386  1.13    mrg </p>
   2387  1.20    mrg <p>Individual checks available in <var class="var">list</var> are &lsquo;<samp class="samp">assert</samp>&rsquo; (compiles libphobos
   2388  1.20    mrg with an extra option <samp class="option">-fassert</samp>).
   2389  1.13    mrg </p>
   2390  1.13    mrg </dd>
   2391  1.20    mrg <dt><code class="code">--with-libphobos-druntime-only</code></dt>
   2392  1.20    mrg <dt><code class="code">--with-libphobos-druntime-only=<var class="var">choice</var></code></dt>
   2393  1.13    mrg <dd><p>Specify whether to build only the core D runtime library (druntime), or both
   2394  1.13    mrg the core and standard library (phobos) into libphobos.  This is useful for
   2395  1.13    mrg targets that have full support in druntime, but no or incomplete support
   2396  1.20    mrg in phobos.  <var class="var">choice</var> can be one of &lsquo;<samp class="samp">auto</samp>&rsquo;, &lsquo;<samp class="samp">yes</samp>&rsquo;, and &lsquo;<samp class="samp">no</samp>&rsquo;
   2397  1.20    mrg where &lsquo;<samp class="samp">auto</samp>&rsquo; is the default.
   2398  1.13    mrg </p>
   2399  1.20    mrg <p>When the option is not specified, the default choice &lsquo;<samp class="samp">auto</samp>&rsquo; means that it
   2400  1.13    mrg is inferred whether the target has support for the phobos standard library.
   2401  1.20    mrg When the option is specified without a <var class="var">choice</var>,  the result is the same as
   2402  1.20    mrg &lsquo;<samp class="samp">--with-libphobos-druntime-only=yes</samp>&rsquo;.
   2403  1.13    mrg </p>
   2404  1.13    mrg </dd>
   2405  1.20    mrg <dt><code class="code">--with-target-system-zlib</code></dt>
   2406  1.20    mrg <dd><p>Use installed &lsquo;<samp class="samp">zlib</samp>&rsquo; rather than that included with GCC.  This needs
   2407  1.12    mrg to be available for each multilib variant, unless configured with
   2408  1.20    mrg <samp class="option">--with-target-system-zlib=&lsquo;<samp class="samp">auto</samp>&rsquo;</samp> in which case the GCC&nbsp;included
   2409  1.20    mrg &lsquo;<samp class="samp">zlib</samp>&rsquo; is only used when the system installed library is not available.
   2410  1.12    mrg </p></dd>
   2411  1.12    mrg </dl>
   2412  1.12    mrg 
   2413  1.10    mrg <hr />
   2414  1.10    mrg <p>
   2415  1.20    mrg <p><a class="uref" href="./index.html">Return to the GCC Installation page</a>
   2416  1.10    mrg </p>
   2417  1.10    mrg 
   2418  1.10    mrg 
   2419  1.10    mrg 
   2420  1.10    mrg 
   2421  1.10    mrg 
   2422  1.10    mrg 
   2423  1.10    mrg 
   2424  1.10    mrg 
   2425  1.10    mrg 
   2426   1.1    mrg 
   2427  1.10    mrg </body>
   2428  1.10    mrg </html>
   2429